diff --git a/18_ADAS-1.ipynb b/18_ADAS-1.ipynb
index cb024c4b..005a5d7d 100644
--- a/18_ADAS-1.ipynb
+++ b/18_ADAS-1.ipynb
@@ -25,7 +25,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
@@ -131,8 +131,7 @@
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
- "hidden": true,
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
"#### 短期総供給曲線(AS曲線)"
@@ -157,14 +156,10 @@
"source": [
"* $Y_t=$ 産出量の対数\n",
"* $Y_{*}=$ 自然産出量の対数\n",
- " * 資本蓄積や技術進歩などにより決定される。\n",
- " * トレンドに対応する。\n",
- " * ADASモデルでは一定と仮定する。下で仮定する需要ショックは一時的なものであり,トレンドには影響しないと仮定する。\n",
+ " * 長期的なトレンドに対応する。\n",
"* $P_t=$ 物価水準の対数\n",
- "* $P_{*}=$ 物価水準の対数\n",
- " * トレンドに対応する。\n",
- " * ここでは一定と仮定する。下で仮定する供給ショックは一時的なものであり,トレンドには影響しないと仮定する。\n",
- "* $v_t=$ 供給ショック"
+ "* $P_t^e=$ $P_t$に対する予測値\n",
+ "* $v_t=$ **一過性の**供給ショック"
]
},
{
@@ -174,24 +169,20 @@
},
"source": [
"式[](eq:18-as)自体はAS曲線の典型的な式となっている。一方で,定量分析のために次の点に関して解釈し直している。\n",
- "* 変数の定義:\n",
- " * 内生変数である$Y_t$と$P_t$は産出量と物価水準の対数としている。同様に,$Y_{*}$,$P_{*}$,$P_t^{e}$も対数としている。対数として解釈することにより,ADASモデルを使ってデータを捉えやすくする利点があるためである。また,より複雑な非線形のADASモデルを均衡周辺でテイラー近似した結果と解釈しても良いだろう。テイラー近似の手法は,研究で広く用いられる方法である。\n",
- "* 産出量の長期的水準:\n",
- " * 長期均衡での産出量は自然産出量$Y_{*}$で与えられるが,これはGDPのトレンドと解釈する。以下の分析では,PHフィルターで計算した値を使うことになる。\n",
- " * 通常のADASモデルの仮定に沿って,総供給ショック$v_t$は$Y_{*}$に影響を与えないとする。即ち,ここでの総供給ショックは,産出量と物価水準がトレンドから乖離する要因となるものであり,その効果は長期的な効果はないと仮定する。言い換えると,トレンドに影響を及ぼす供給ショックもあり得る訳だが,それらは対象とはなっていないことになる。\n",
- "* 物価水準の長期的水準:\n",
- " * ADASモデルでの長期的な物価水準とは$Y_t=Y_{*}$が満たされる$P_t$であり,基本的にはどのような値をとっても構わない。例えば,総供給ショック$v_t$が`0`から`0.5`に増加(不利な総供給ショック)し,`0.5`が永続的に続く場合,長期的な物価水準は永続的に高くなる。しかし,ここで考える総供給ショック$v_t$は一過性のショックと仮定する。即ち,ある時は`0.3`や`-0.1`などの値を取るが,それは短期的な現象であり,長期的には`0`に戻ると仮定する。この仮定により,長期的な物価水準$P_{*}$が存在することになり,それを物価水準のトレンドだと解釈する。産出量と同様に,HPフィルターを使って物価水準の長期的なトレンドを算出することになる。"
+ "* 対数の役割:\n",
+ " * 内生変数である$Y_t$と$P_t$及び定数の$Y_{*}$は産出量と物価水準の対数としている。対数として解釈することにより,ADASモデルを使ってデータを捉えやすくする利点がある。また,より複雑な非線形のADASモデルを均衡周辺でテイラー近似した結果と解釈しても良いだろう。テイラー近似の手法は,研究で広く用いられる方法である。\n",
+ "* 供給ショックの種類:\n",
+ " * 一過性の総供給ショック$v_t$は長期的なトレンドから**一時的な**乖離を引き起こす要因となるものである。後に導入する物価水準の長期的トレンド$P_{*}$には影響を及ぼさないと仮定する。もちろん,供給ショックとなるため$Y_{*}$にも影響しない。"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
- "hidden": true,
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
- "#### 総需要曲線(AD曲線)"
+ "#### 短期総需要曲線(AD曲線)"
]
},
{
@@ -204,7 +195,7 @@
"Y_t=b-c P_t + u_t, \\qquad b,c>0\n",
"$$(eq:18-ad)\n",
"\n",
- "* $u_t=$ 総需要ショック"
+ "* $u_t=$ **一過性の**需要ショック"
]
},
{
@@ -214,8 +205,9 @@
},
"source": [
"AD曲線はIS-LMモデルから導出することができるが,その場合,政府支出やマネーストックなど様々な変数に依存した複雑な式となる。式[](eq:18-ad)は対数線形の簡単な式となっているが,その裏にあるのが次の考え方である。\n",
- "* 政府支出や租税,マネーストック,更には,消費や投資水準はパラメータである$b$と$c$に含まれていると考える。また,それらの変化は全て総需要ショック$u_t$が捉えていると仮定する。\n",
- "* 総供給ショック$v_t$と同様に,総需要ショック$u_t$は長期的な産出量水準には影響は与えないと仮定を置くとともに,ショック自体も永続的に維持されるのではなく,時間とともにフェードアウトする短期的ショックと仮定する。\n",
+ "* 政府支出や租税,マネーストック,更には,消費や投資水準は$b$及び$u_t$に含まれていると考える。また,それらの変化は全て需要ショックである$u_t$が捉えていると仮定する。\n",
+ "* 需要ショックには次の2つがあると仮定する。\n",
+ " * 需要ショック$u_t$は自然産出量水準$Y_t^{*}$には影響は与えないと仮定を置くとともに,ショック自体も永続的に維持されるのではなく,時間とともにフェードアウトする**一過性の**ショックと仮定する。\n",
"\n",
"このように簡略化することにより,景気循環データをADASモデルの枠組みの中で扱いやすくなる利点がある。"
]
@@ -224,8 +216,7 @@
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
- "hidden": true,
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
"#### 適応的期待"
@@ -241,15 +232,14 @@
"P_t^e= P_{t-1}\n",
"$$(eq:18-pe)\n",
"\n",
- "適応的期待は「後ろ向き」の期待形成を仮定することになり,批判の的にもなり易い。しかし,最前線の研究でも取り上げられる場合があり,データの特徴を捉えた仮定と言えるだろう。"
+ "適応的期待は「後ろ向き」の期待形成を仮定することになり,批判の的にもなり易い。しかし,最前線の研究でも取り上げられ,データの特徴を捉えた仮定と言えるだろう。"
]
},
{
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
- "hidden": true,
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
"#### 総需要・総供給ショック"
@@ -261,39 +251,20 @@
"hidden": true
},
"source": [
- "既に触れたが,ショック項$u_t$と$v_t$は次の2点を満たすと仮定する。\n",
- "* 自然産出量$Y_{*}$に影響を及ぼさない。\n",
+ "既に触れたが,一過性のショック項$u_t$と$v_t$は次の2点を満たすと仮定する。\n",
+ "* 物価水準トレンド$P_{*}$と自然産出量$Y_{*}$に影響を及ぼさない。\n",
"* 短期的・一時的なショックであり,永続的に正または負の一定の値を取り続けない。\n",
"\n",
- "一方で,確率的シミュレーションを行う際は,毎期毎期において総需要ショックと総供給ショックが発生する状況を考える。その場合,次の2つの仮定の内の一つを採用することになる。\n",
- "1. ホワイト・ノイズ\n",
+ "一方で,確率的シミュレーションを行う際は,毎期毎期において一過性の総需要ショックと総供給ショックが発生する状況を考える。その場合,次のホワイト・ノイズの仮定を採用することになる。\n",
"\n",
"$$\n",
"\\begin{aligned}\n",
"u_t&\\sim\\text{N}(0,\\sigma_u^2)\\\\\n",
"v_t&\\sim\\text{N}(0,\\sigma_v^2)\n",
"\\end{aligned}\n",
- "$$\n",
- "\n",
- "2. AR(1)自己回帰モデル\n",
- "\n",
- "$$\n",
- "\\begin{aligned}\n",
- "u_t&=\\rho_u u_{t-1}+e_{ut},\\qquad u_t\\sim\\text{N}(0,\\sigma_{eu}^2)\\\\\n",
- "v_t&=\\rho_v v_{t-1}+e_{vt},\\qquad v_t\\sim\\text{N}(0,\\sigma_{ev}^2)\n",
- "\\end{aligned}\n",
"$$"
]
},
- {
- "cell_type": "markdown",
- "metadata": {
- "hidden": true
- },
- "source": [
- "この章でおこなう確率的シミュレーションではホワイト・ノイズを仮定するが,AR(1)を使う場合は付録で扱うことにする。"
- ]
- },
{
"cell_type": "markdown",
"metadata": {
@@ -310,7 +281,7 @@
"\n",
"このようなショックを「構造ショック」と呼ぶ。なぜそう呼ばれるかを直感的に考えてみよう。経済メカニズムにより発生する変数の動きは,時間的なラグが存在したり(例えば,雇用の調整にはコストが掛かるが,それが非常に高い失業率の持続性の一つの要因と考えられる),他の変数の過去の値から影響を受けたりする(例えば,失業している間に労働者のスキルは失われ,それが将来のGDPに負の影響を及ぼす可能性がある)。このような経済メカニズムに起因する効果を捉えたのが(ii)の過去の変数である。2変数の時間的なラグが十分に長くなれば,殆どの経済メカニズムの効果は捉えることができるため,残っているのは(iii)にある経済メカニズムと無関係な純粋にランダムはショックということになる。それを「構造ショック」としてGDPショック,失業率ショックと呼んでいる。\n",
"\n",
- "ADASモデルの$u_t$と$v_t$はどうかというと,ADASモデルには十分な変数のラグが無いため上の「構造ショック」の3つの特徴を必ずしも満たすとは限らない。むしろ,$u_t$には$p_t$で捉えられていない政府,中央銀行,企業や消費者の$y_t$($Y_t$のトレンドからの乖離)に対する影響を全て反映しており,ここではそれを総需要ショックと呼んでいる。同様に,$v_t$は適応的期待と$y_t$で捉えられていない,企業などの$p_t$($P_t$のトレンドからの乖離)に対する影響を総需要ショックと呼んでいる。即ち,$u_t$と$v_t$には経済メカニズムにより発生するショックを含んでいると考えることができる。従って,$u_t$と$v_t$は,VARモデルでいう構造ショックとは異なることになる。\n",
+ "ADASモデルの$u_t$と$v_t$はどうかというと,ADASモデルには十分な変数のラグが無いため上の「構造ショック」の3つの特徴を必ずしも満たすとは言えない。むしろ,$u_t$には$p_t$で捉えられていない政府,中央銀行,企業や消費者の$y_t$($Y_t$のトレンドからの乖離)に対する影響を全て反映しており,ここではそれを総需要ショックと呼んでいる。同様に,$v_t$は適応的期待と$y_t$で捉えられていない,企業などの$p_t$($P_t$のトレンドからの乖離)に対する影響を総需要ショックと呼んでいる。即ち,$u_t$と$v_t$には経済メカニズムにより発生するショックを含んでいると考えることができる。従って,$u_t$と$v_t$は,VARモデルでいう構造ショックとは異なることになる。\n",
"```"
]
},
@@ -318,11 +289,10 @@
"cell_type": "markdown",
"metadata": {
"heading_collapsed": true,
- "hidden": true,
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
- "#### 長期均衡"
+ "#### 長期均衡とトレンドの決定"
]
},
{
@@ -332,9 +302,23 @@
},
"source": [
"長期均衡または定常状態(steady state)では次の条件が満たされることになる。\n",
- "* $u_{t}=v_t=0$:ショックはない\n",
- "* $P_t=P_*$\n",
- "* $Y_t=Y_*$"
+ "* $u_{t}=v_t=0$\n",
+ "* $P_t=P_{t-1}=P_{*}$\n",
+ "* $Y_t=Y_{*}$\n",
+ "\n",
+ "この場合AD曲線[](eq:18-ad)は次の様になる。\n",
+ "\n",
+ "$$\n",
+ "Y_{*}=b-c P_{*}\n",
+ "$$(eq:18-lrad)\n",
+ "\n",
+ "この式は自然産出量と物価水準の長期的な関係を表している。\n",
+ "もちろん,トレンドのデータはある程度の変動を示しながら持続的に増加しており,式[](eq:18-lrad)は満たされない。\n",
+ "ここでは,数式を簡単にしつつデータと整合性がある様にするための単なる仮定であり,これにより以下で説明する式[](eq:18-as_small)と式[](eq:18-ad_small)を簡単に導出することが可能となる。\n",
+ "\n",
+ "一方で,自然産出量$Y_{t*}$と物価水準のトレンド$P_{t*}$が増加する改良バージョンを付録Aで展開している。\n",
+ "こちらの方がよりデータと整合的ではあるが,結局は,式[](eq:18-as_small)と式[](eq:18-ad_small)が導出されることになる。\n",
+ "わざわざモデルを複雑にしなくても同じ均衡式が導出できるのであれば,簡単なバージョンで十分である。Simplicity is the best approach!"
]
},
{
@@ -354,9 +338,9 @@
"hidden": true
},
"source": [
- "ADASモデルを図示すると{numref}`fig:18-adas0`のようになる。長期均衡においてAD曲線とAS曲線は$P_*$と$Y_*$で交差することになる。短期均衡を考えるために,総供給ショック$v_t$が1期間だけ正の値を取り,その後は`0`になるとしよう。ショックが発生するとAS曲線は上方シフトし,産出量は減少し物価水準は上昇する。適応的期待により,ショックがなくなってもAS曲線は直ぐには長期均衡に戻らず,徐々に下方シフトすることになり,最終的には長期均衡に経済は戻ることになる。\n",
+ "ADASモデルを図示すると{numref}`fig:18-adas0`のようになる。長期均衡においてAD曲線とAS曲線は$P_*$と$Y_*$で交差することになる。短期均衡を考えるために,供給ショック$v_t$が`1`期間だけ正の値を取り,その後は`0`になるとしよう。ショックが発生するとAS曲線は上方シフトし,産出量は減少し物価水準は上昇する。適応的期待により,ショックがなくなってもAS曲線は直ぐには長期均衡に戻らず,徐々に下方シフトすることになり,最終的には長期均衡に経済は戻ることになる。\n",
"\n",
- "このような分析は,教科書で解説される典型的な例である。長期均衡は安定的であり,産出量と物価水準は少しずつ動くという結果は直感的であり,わかり易い。しかし,このような定性的な分析からは,ショック発生後に経済はどれだけの時間をかけて長期均衡に戻るかについては何もわからない。また,長期均衡に戻るスピードが何に依存しているのかも分からない。AD曲線とAS曲線の傾きが影響を及ぼしそうだと想像できるが,傾きが急だと戻るスピードが早いのだろうか,遅いのだろうか。このような問は,定量分析をおこなうことにより簡単に確認することが可能となる。"
+ "このような分析は,教科書で解説される典型的な例である。長期均衡は安定的であり,産出量と物価水準は少しずつ動くという結果は直観的であり分かり易い。しかし,このような定性的な分析からは,ショック発生後に経済はどれだけの時間をかけて長期均衡に戻るかについては何もわからない。また,長期均衡に戻るスピードが何に依存しているのかも分からない。AD曲線とAS曲線の傾きが影響を及ぼしそうだと想像できるが,傾きが急だと戻るスピードが早いのだろうか,遅いのだろうか。このような問は,定量分析をおこなうことにより簡単に確認することが可能となる。"
]
},
{
@@ -391,9 +375,9 @@
"hidden": true
},
"source": [
- "ADASモデルとデータと整合性があるようにするために次の変数を定義しよう。\n",
- "* $p_t=P_t-P_*$:価格水準のトレンドからの乖離率\n",
- "* $y_t=Y_t-Y_*$:産出量のトレンドからの乖離率\n",
+ "ADASモデルとデータの親和正を高めるために次の変数を定義しよう。\n",
+ "* $p_t=P_t-P_{*}$:価格水準のトレンドからの乖離率\n",
+ "* $y_t=Y_t-Y_{*}$:産出量のトレンドからの乖離率\n",
"\n",
"$p_t$と$y_t$を使ってAS曲線とAD曲線を書き換えることにする。"
]
@@ -439,10 +423,12 @@
"hidden": true
},
"source": [
- "この式はAS曲線を産出量と物価水準のトレンドからの乖離率で表している。また,この式から次のことが言える。\n",
- "* $p_{t-1}$と$y_t$を所与とすると,総供給ショック$v_t$の1単位は物価水準のトレンドからの乖離を1%発生させる。\n",
+ "この式はAS曲線を産出量と物価水準のトレンドからの乖離率で表している。また,この式から供給ショック$v_t$について次のことが言える。\n",
+ "\n",
+ "* 直接効果:$p_{t-1}$と$y_t$を所与とすると,$v_t$の1単位上昇は物価水準のトレンドからの乖離を1%発生させる。\n",
+ "* 間接効果:$p_{t}$と$y_t$は均衡で決定されるため,$v_t$が変化すると$y_t$も変化し,その効果を通して$p_{t}$が変化する。\n",
"\n",
- "(理由)$p_{t-1}$と$y_t$を一定として$v_t=1$とすると,$p_t=1$となる。"
+ "直接効果と間接効果の相互作用により供給ショック$v_t$は$p_t$に影響を及ぼすことになる。"
]
},
{
@@ -461,46 +447,32 @@
"hidden": true
},
"source": [
- "まず長期均衡でのAD曲線を考えてみよう。\n",
- "\n",
- "$$\n",
- "Y_{*}=b-c P_{*} \n",
- "$$ (eq:18-adss)\n",
- "\n",
- "$u_t=0$となるため,上の式が成立することになる。"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "hidden": true
- },
- "source": [
- "次に,式[](eq:18-ad)の両辺から$Y_{*}$を引くと次式となる。\n",
- "\n",
- "$$Y_t-Y_*=b-Y_*-c P_t + u_t$$\n",
- "\n",
- "また左辺の$b-Y_{*}$を削除するために式[](eq:18-adss)を使うと\n",
+ "式[](eq:18-ad)の両辺から式[](eq:18-lrad)を引くと次の1行目となる。\n",
"\n",
"$$\n",
- "Y_t-Y_*=cP_*-c P_t + u_t\n",
+ "Y_t-Y_{*}=(b-b)-c(P_t-P_{*})+ u_t\n",
"$$\n",
"\n",
- "となり,上の定義を使い次式を導出できる。\n",
- "\n",
"$$\n",
- "y_t=-cp_t + u_t\n",
+ "\\begin{aligned}\n",
+ "&\\Downarrow\\quad\\text{上の定義を使う}\\\\\n",
+ "\\\\\n",
+ "y_t &= -cp_{t} +v_t\n",
+ "\\end{aligned}\n",
"$$ (eq:18-ad_small)"
]
},
{
"cell_type": "markdown",
- "metadata": {
- "hidden": true
- },
+ "metadata": {},
"source": [
- "この式はAD曲線を産出量と物価水準のトレンドからの乖離率で表している。また,この式から次のことが言える。\n",
- "* $p_{t}$を所与とすると,総需要ショック$u_t$の1単位はGDPのトレンドからの乖離を1%発生させる。"
+ "この式はAS曲線を産出量と物価水準のトレンドからの乖離率で表している。また,この式から供給ショック$v_t$について次のことが言える。\n",
+ "\n",
+ "この式はAD曲線を産出量と物価水準のトレンドからの乖離率で表している。また,この式から需要ショック$u_t$について次のことが言える。\n",
+ "* 直接効果:$p_{t}$を所与とすると,$u_t$の1単位上昇は産出量のトレンドからの乖離を1%発生させる。\n",
+ "* 間接効果:$p_{t}$と$y_t$は均衡で決定されるため,$u_t$が変化すると$p_t$も変化し,その効果を通して$y_{t}$が変化する。\n",
+ "\n",
+ "直接効果と間接効果の相互作用により供給ショック$u_t$は$y_t$に影響を及ぼすことになる。"
]
},
{
@@ -573,7 +545,9 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "jp-MarkdownHeadingCollapsed": true
+ },
"source": [
"### トレンドからの乖離率"
]
@@ -584,12 +558,12 @@
"hidden": true
},
"source": [
- "この章では式[](eq:18-as_small)と式[](eq:18-ad_small)を使い定量分析をおこなうが,まず,それらの式と整合性があるデータの特徴を考える。$y_t$と$p_t$はトレンドからの乖離率となるため,`py4macro`の`trend()`関数を使い,対応する変数を作成することにする。"
+ "ここでは式[](eq:18-as_small)と式[](eq:18-ad_small)を使い定量分析をおこなうが,まず,それらの式と整合性があるデータの特徴を考える。$y_t$と$p_t$はトレンドからの乖離率となるため,`py4macro`の`trend()`関数を使い,対応する変数を作成することにする。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 2,
"metadata": {
"hidden": true
},
@@ -600,11 +574,25 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 3,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "Index(['gdp', 'consumption', 'investment', 'government', 'exports', 'imports',\n",
+ " 'capital', 'employed', 'unemployed', 'unemployment_rate', 'hours',\n",
+ " 'total_hours', 'inflation', 'price', 'deflator'],\n",
+ " dtype='object')"
+ ]
+ },
+ "execution_count": 3,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"df.columns"
]
@@ -622,11 +610,215 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 4,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "
\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " gdp | \n",
+ " consumption | \n",
+ " investment | \n",
+ " government | \n",
+ " exports | \n",
+ " imports | \n",
+ " capital | \n",
+ " employed | \n",
+ " unemployed | \n",
+ " unemployment_rate | \n",
+ " hours | \n",
+ " total_hours | \n",
+ " inflation | \n",
+ " price | \n",
+ " deflator | \n",
+ " gdp_cycle | \n",
+ " deflator_cycle | \n",
+ "
\n",
+ " \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 1980-01-01 | \n",
+ " 269747.5 | \n",
+ " 153290.7 | \n",
+ " 65029.2 | \n",
+ " 73039.5 | \n",
+ " 18383.8 | \n",
+ " 24278.8 | \n",
+ " 833041.425949 | \n",
+ " 5506.000000 | \n",
+ " 107.666667 | \n",
+ " 1.900000 | \n",
+ " 124.7 | \n",
+ " 686598.200000 | \n",
+ " 5.766667 | \n",
+ " 71.011939 | \n",
+ " 91.2 | \n",
+ " 0.003269 | \n",
+ " -0.025798 | \n",
+ "
\n",
+ " \n",
+ " 1980-04-01 | \n",
+ " 268521.8 | \n",
+ " 153551.9 | \n",
+ " 65316.6 | \n",
+ " 72164.5 | \n",
+ " 18631.4 | \n",
+ " 25454.5 | \n",
+ " 843748.581006 | \n",
+ " 5525.666667 | \n",
+ " 110.000000 | \n",
+ " 1.966667 | \n",
+ " 124.8 | \n",
+ " 689603.200000 | \n",
+ " 8.166667 | \n",
+ " 72.917251 | \n",
+ " 93.4 | \n",
+ " -0.010845 | \n",
+ " -0.006555 | \n",
+ "
\n",
+ " \n",
+ " 1980-07-01 | \n",
+ " 274183.1 | \n",
+ " 155580.0 | \n",
+ " 65765.9 | \n",
+ " 72663.8 | \n",
+ " 18449.3 | \n",
+ " 23885.7 | \n",
+ " 855449.381902 | \n",
+ " 5561.333333 | \n",
+ " 116.000000 | \n",
+ " 2.033333 | \n",
+ " 124.0 | \n",
+ " 689605.333333 | \n",
+ " 8.200000 | \n",
+ " 73.897654 | \n",
+ " 94.4 | \n",
+ " 0.000457 | \n",
+ " -0.000482 | \n",
+ "
\n",
+ " \n",
+ " 1980-10-01 | \n",
+ " 279601.8 | \n",
+ " 156162.4 | \n",
+ " 66017.5 | \n",
+ " 74761.1 | \n",
+ " 19705.4 | \n",
+ " 23716.5 | \n",
+ " 867991.306098 | \n",
+ " 5551.333333 | \n",
+ " 123.333333 | \n",
+ " 2.166667 | \n",
+ " 124.0 | \n",
+ " 688365.333333 | \n",
+ " 8.100000 | \n",
+ " 74.767147 | \n",
+ " 95.4 | \n",
+ " 0.010468 | \n",
+ " 0.005515 | \n",
+ "
\n",
+ " \n",
+ " 1981-01-01 | \n",
+ " 281995.7 | \n",
+ " 156757.7 | \n",
+ " 66259.0 | \n",
+ " 76127.6 | \n",
+ " 20289.5 | \n",
+ " 24174.1 | \n",
+ " 878387.487376 | \n",
+ " 5568.666667 | \n",
+ " 124.333333 | \n",
+ " 2.200000 | \n",
+ " 123.6 | \n",
+ " 688287.200000 | \n",
+ " 6.833333 | \n",
+ " 75.690241 | \n",
+ " 95.6 | \n",
+ " 0.009441 | \n",
+ " 0.003126 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ],
+ "text/plain": [
+ " gdp consumption investment government exports imports \\\n",
+ " \n",
+ "1980-01-01 269747.5 153290.7 65029.2 73039.5 18383.8 24278.8 \n",
+ "1980-04-01 268521.8 153551.9 65316.6 72164.5 18631.4 25454.5 \n",
+ "1980-07-01 274183.1 155580.0 65765.9 72663.8 18449.3 23885.7 \n",
+ "1980-10-01 279601.8 156162.4 66017.5 74761.1 19705.4 23716.5 \n",
+ "1981-01-01 281995.7 156757.7 66259.0 76127.6 20289.5 24174.1 \n",
+ "\n",
+ " capital employed unemployed unemployment_rate hours \\\n",
+ " \n",
+ "1980-01-01 833041.425949 5506.000000 107.666667 1.900000 124.7 \n",
+ "1980-04-01 843748.581006 5525.666667 110.000000 1.966667 124.8 \n",
+ "1980-07-01 855449.381902 5561.333333 116.000000 2.033333 124.0 \n",
+ "1980-10-01 867991.306098 5551.333333 123.333333 2.166667 124.0 \n",
+ "1981-01-01 878387.487376 5568.666667 124.333333 2.200000 123.6 \n",
+ "\n",
+ " total_hours inflation price deflator gdp_cycle \\\n",
+ " \n",
+ "1980-01-01 686598.200000 5.766667 71.011939 91.2 0.003269 \n",
+ "1980-04-01 689603.200000 8.166667 72.917251 93.4 -0.010845 \n",
+ "1980-07-01 689605.333333 8.200000 73.897654 94.4 0.000457 \n",
+ "1980-10-01 688365.333333 8.100000 74.767147 95.4 0.010468 \n",
+ "1981-01-01 688287.200000 6.833333 75.690241 95.6 0.009441 \n",
+ "\n",
+ " deflator_cycle \n",
+ " \n",
+ "1980-01-01 -0.025798 \n",
+ "1980-04-01 -0.006555 \n",
+ "1980-07-01 -0.000482 \n",
+ "1980-10-01 0.005515 \n",
+ "1981-01-01 0.003126 "
+ ]
+ },
+ "execution_count": 4,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"for c in ['gdp','deflator']:\n",
" \n",
@@ -646,11 +838,22 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 5,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"ax_ = df.plot(y=['gdp_cycle','deflator_cycle'])\n",
"ax_.axhline(0, color='red')\n",
@@ -679,11 +882,20 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 6,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "GDPのトレンドからの乖離率の標準偏差:0.014833\n",
+ "デフレータのトレンドからの乖離率の標準偏差:0.007490\n"
+ ]
+ }
+ ],
"source": [
"y_std = df.loc[:,'gdp_cycle'].std()\n",
"p_std = df.loc[:,'deflator_cycle'].std()\n",
@@ -708,6 +920,7 @@
"editable": true,
"heading_collapsed": true,
"hidden": true,
+ "jp-MarkdownHeadingCollapsed": true,
"slideshow": {
"slide_type": ""
},
@@ -719,11 +932,20 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 7,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "GDPのトレンドからの乖離率の自己相関係数:0.697\n",
+ "デフレータのトレンドからの乖離率の自己相関係数:0.830\n"
+ ]
+ }
+ ],
"source": [
"y_autocorr = df.loc[:,'gdp_cycle'].autocorr()\n",
"p_autocorr = df.loc[:,'deflator_cycle'].autocorr()\n",
@@ -746,6 +968,7 @@
"editable": true,
"heading_collapsed": true,
"hidden": true,
+ "jp-MarkdownHeadingCollapsed": true,
"slideshow": {
"slide_type": ""
},
@@ -757,7 +980,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 8,
"metadata": {
"editable": true,
"hidden": true,
@@ -766,7 +989,15 @@
},
"tags": []
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "GDPとデフレータの乖離率の相関係数:-0.153\n"
+ ]
+ }
+ ],
"source": [
"yp_corr = df.loc[:,['gdp_cycle', 'deflator_cycle']].corr().iloc[0,1]\n",
"print(f'GDPとデフレータの乖離率の相関係数:{yp_corr:.3f}')"
@@ -774,7 +1005,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 9,
"metadata": {
"editable": true,
"slideshow": {
@@ -784,7 +1015,42 @@
"remove-cell"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "application/papermill.record/text/plain": "6"
+ },
+ "metadata": {
+ "scrapbook": {
+ "mime_prefix": "application/papermill.record/",
+ "name": "t_shift"
+ }
+ },
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "application/papermill.record/text/plain": "0.433"
+ },
+ "metadata": {
+ "scrapbook": {
+ "mime_prefix": "application/papermill.record/",
+ "name": "yp_corr_shift"
+ }
+ },
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAGcCAYAAAAh7lEjAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAADA6UlEQVR4nOydd5gb5fW2n1HX9t6bezeuNGNj02xj01sgPwgtISEk2ISYkABfaIFgCEkIBJIACYTqQDABm+JCtQH33u319t61qz7fH++8M6O6klZbtDr3de2l3dFImtVqZ877nOecI4iiKIIgCIIgCCLG0Qz2ARAEQRAEQUQDCmoIgiAIghgWUFBDEARBEMSwgIIagiAIgiCGBRTUEARBEAQxLKCghiAIgiCIYQEFNQRBEARBDAsoqCEIgiAIYligG+wDGCjcbjdqamqQnJwMQRAG+3AIgiAIgggBURTR2dmJgoICaDTBtZi4CWpqampQXFw82IdBEARBEEQEVFZWoqioKOg+cRPUJCcnA2BvSkpKyiAfDUEQBEEQodDR0YHi4mL5Oh6MuAlqeMopJSWFghqCIAiCiDFCsY6QUZggCIIgiGEBBTUEQRAEQQwLKKghCIIgCGJYQEENQRAEQRDDAgpqCIIgCIIYFlBQQxAEQRDEsICCGoIgCIIghgUU1BAEQRAEMSygoIYgCIIgiGEBBTUEQRAEQQwLKKghCIIgCGJYQEENQRAEQRDDAgpqCIIghiFdNide+PwYmrtsg30oBDFgUFBDEAQxDPnrZ0fx2NqD+NP6I4N9KAQxYFBQQxAEMQz57kQLAGBXVfsgHwlBDBwU1BAEQQwzHC43dkvBzKG6Drjc4iAfEUEMDBTUEARBDDMO1HbA5nQDAKwON040WQb5iAhiYKCghiAIYpix/WSrx88HajsG6UgIYmChoIYgCGKYsb2iDQCg1QgAgP0U1BBxAgU1BEEQw4ztFUypOX9CLgBSaoj4gYIagiCIYURDpxVVrT0QBODa00oAUFBDxA8U1BAEQQwjdkipp7E5yZhVmg5BAOo7bNSEj4gLKKghYp5HP9yPK/+6CVaHa7APhSAGHZ56mlGahkSjDmWZiQCAA7Wdg3lYBDEgUFBDxDxvfFeJrSdbsa+GmowR8YPT5YYo+vaf2XGyDQAwvTgdADAhPxkAsL+W/j+I4Y9usA+AIPqC1eFCl80JAKhrJ3mdGN68taUCL39djsZOG1q67Rifl4L375gDvZatTx0uN3ZXtwFgSg0ATMhLwZo9daTUEHEBKTVETNNiscvf17b3DOKREET/84dPD+NgXSeaLXaIIjMAH65XgpUDtR2wOtxIMekwMisJADCxIEW+jyCGOxTUEDGNOqip77AO4pEQRP/SYXWgvoOpkf+9/UzMLGXppf01SrCyq7INADCtJB0aqUfNhHwW1Bxt6ILNSb4zYnhDQQ0R0zR7KDUU1BDDl6MNXQCA3BQjppek45SiNACejfX4vKdTilLlbfmpJqQl6OF0izhS3zVwB0wQgwAFNURM02JRfDR1FNQQwxge1IzO8UwrqZWaPdUsqJlSqAQ1giBgQh6loIj4gIIaIqZp7lKUmjpKPxHDmGM8qMlmQY1S1dQBURTRY3fJ/pqpkorDKc4wAwAaOslMTwxvqPqJiGmavTw1brcoewkIYjjhrdSMyUmGXiug0+pEVWsP+/yLQE6yEXmpJo/HZiYZAQBN1ICPGOaQUkPENC0qpcbhEtHSbQ+yN0HELkcbWVAzSgpqDDoNRucwteZAbYfsp5mq8tNwMhMNADyVTYIYjlBQQ8Q0aqUGIF8NMTyxOlyobOkGoCg1ADBRqmzaX9uh8tOk+Tw+M4kFNS0WCmqI4Q2ln4iYptniKafXtlsxudB3pRqv7K1ux9dHm1DbbkVjpw1Lpubjwin5g31YRJicaLLALQIpJh2ypVQSwMzC72xnZuFjkpLjX6mh9BMRH1BQQ8Q0fOWZnqBHa7fDr1l428lW3P7aNtxxzhhcf3rpQB/ioGF1uHD1C5vRbVd6k+ysbKOgJgZR+2kEQfGMcaVme0WrrFpO8RfUSEqNt7JJEMMNSj8RMQ331EwqYCfyOq+uwi63iN/8dw/qO2x4f2f1gB/fYFLd1oNuuwtGnQY3nlkGgFWIud2+84KIoY23SZjDg5qmLtZhuDDNjCyVksPhSk2LxU5/f2JYQ0ENEbPYnC50SnOfJkk9O7znP721pRIH61iZ64mm7oE9wEGG+4uKMxJw35IJEAQW5NFqPfbgJmHvoCY1QY/CNLP885QAqdcMySjscovosDr66SgJYvDpt6Dmn//8JyZPnoyioiLMnj0bX331VcB9q6urcc0116CsrAyFhYVYvnw5bDbPi9PXX3+NefPmoaioCCNGjMCdd94Ji8XSX4dPxAA89aTTCBiby6pA6joUpabD6sBTnxySf27qsqEzjk7ovMNyfqoJOq1GXsE3dJKZOtY4FkCpAZQmfID/1BPAKqVSTMxt0EQVUMQwpl+CmldffRX33nsvVq1ahaqqKqxYsQJLlizB8ePHffa12+04//zzUVRUhKNHj2Lfvn3Ytm0bli9fLu9z8OBBLFmyBHfeeSeqqqqwY8cO7NmzB8uWLeuPwydiBF6emp5oQH4a68uhHpXw7IajaLbYMTI7UV6plseRWsNTcXkp7L3JSZaCmg4yi8YSLreI401sATc6O9nnfp6CAvybhDm8V00zmYWJgaR6G2BpHrCX65eg5sEHH8Tdd9+NCRMmAACuuuoqzJs3D3/5y1989n377bdRX1+Pxx57DDqdDmlpaXj66afx4osvoqmpCQDw3Xff4aabbsIVV1wBAEhLS8Py5cvxzjvv9MfhEzECV2oyEw3yhbuu3QpRFFHT1oOXvy4HANy/ZCJGZScCAI43xc/sG1mpkdITudJ7REpNbFHZ0g270w2jToPCdLPP/RPUQY2fcm6O3KuG0o/EQCGKwFs3ACtHARXfDMhLRj2oqaiowLFjx3DRRRd5bL/ooouwdu1an/03bNiAhQsXwmAwyNtmzpyJzMxMrF+/HgBwww034Omnn/Z43J49e5CSkoJA2Gw2dHR0eHwRwws5qEkyyB1Uu+3MZ/PR3jrYXW7MLE3H/HHZKMtkQU08KTXq9BOgKDX1pNTEFNwkPDI7CVo/3bJnl6UjLUGPM0ZmIjVBH/B5qAIqPhFFEX//4jg+2F0z8C/echzoqAK0eiBv6oC8ZNRLumtq2BtXUFDgsb2wsBDV1b7VJzU1NZg8ebLP9kD7A8Arr7yCBx98EC+88ELA43jsscfw4IMPhnPoRIzBe25kJBqRYNAh1axHe48D9e1WrD9YDwBYPDkPgiBghKTUlDfHjw+LBzV5XkENKTWxRSCTMCczyYiv7jkHul7Gg2QkUvopHvnscCMeXXMAKSYdlk4t6P0B0eT4RnZbfBpgSBiQl4y6UqPXs5WCRuP51OreCt77e+8baP+enh7cfPPNuPPOO/Hqq6/ixhtvDHgc9957L9rb2+WvysrKMH4LIhZQp58AxTtyuL4L3x5vAQCcNyEXADAik6ef4ieo4Z4aWamR3h9SamIL70GW/kgy6mDSa4M+T1YSjUqIR1766gQAoMPqhCgOcDn/8c/Y7cizB+wlo67UFBUVAWAKzOjRo+XttbW1KCws9Ls/V3fUeO/f3NyMhQsXIjs7G/v27fNRgrwxGo0wGn37NRDDBx7UcBNwXqoJh+o78dbWSjjdIkZlJ6IsiwUzslITJ0GN1eFCazer9MpPYT4MRamhoCaW4A0l/flpwoEH/zQqIX44XN+JL480yT/bXW4YdcGD36jhdgEnvmDfj1wQ0kNEUURTlx3ZyZFfu6Ou1OTm5mLatGlYs2aNx/ZPP/0Uixcv9tl/0aJF+OSTT+B0OuVtBw8eRENDA84991wAgMPhwNKlS3HWWWdhzZo1vQY0RHzQrPLUAIpS88XhRgCKSgMApRksqGnvcaA1Dk7qvEeNWa9FipmtXWSjsJ+uy8TQpU0KTjMSA/tlQoEmdccfL399wuNnq8M9cC9euxOwtgPGVCB/WkgPef7z45j96Dp8tLc24pftl+qnFStW4IknnsDhw4cBAKtXr8batWtx++23++y7ZMkS5OTk4P7774fL5UJ7ezvuuOMO3HTTTcjKygIAPP3000hISMDTTz8dMI1FxB/cG5CpUmrUnKsKaswGrZyGiYcUVA1PPaWZ5P+ZnBR2UWvstFFX2RiCKytpCYZe9gwOVT/FNm63iFc2l+NAbWhFLy0WO97d7ulLtTldAfbuB45/zm5HzAW0oSWFvj3BSr/5xPlI6Jeg5tprr8X999+PpUuXoqCgAI8++ig++OADjB49GlVVVSgqKsKqVasAADqdDh999BH279+P4uJiTJo0CVOmTMGf/vQn+fnWrl2Lbdu2obi4GEVFRR5fmzdv7o9fgYgBlPQTu1jnq4KatAQ9ZpSkeew/Iit+UlB1XpVPAJCVZIQgAE63iJZuurDFCm3S3yqjr0EN9amJadbsrcUDq/fhV+/uCWn/1789CZvTjcmFKTDp2aXeNpBKDffTjAjdT8Mn0bf24fzUbwMtb7vtNtx2220+24uKilBVVeWzbfXq1QGfa+PGjVE/PiL2afby1OSqLuALxuVAp/WM2cuyErHpWDNOxEFQI1c+pSg+DL1Wg8xEA5q67GjosPmdEUQMLWxOFyzSQNL0Pgc17PFtPQ44XW6f/w9iaPPZIZZWP1Db0evfTxRFvPZtBQDglrNG4MH/7YfV4R44pcbRo/SlGTk/pIeIooiqVqYw98X3RZ9qIiaxO93otDIfFq/qUKsSaj8NZ6Sk1JyQyro/3V+Pn762fVjOwvGn1ABATrJUAUVl3TEB99NoNQKSTX1bg6YnGCAIrB8aN5ETsYEoivjyCAtq7E43ypuD99uqau1BbbsVeq2AxZPzYZLMwQPmqan4BnDZgOQCIGtMSA9p7LTB5mTH12qJ/PNJQQ0Rk/BIXqsRkGJiBsrCNDPMei0SDFrMG5vl8xilAZ8FLRY77nprJz7cU4u1eyI3pQ1VvHvUcGRfDZV1xwSyn8ashyZQHxqXA/jgLmDLP4I+l1YjyGpPs4X+/rHE4fouj1YMh+s7g+6/s7INAOs2bdJrYeTpp4FSauRS7vlAiD7YylYlUOvL57Pf0k8E0Z/wD316gkE+2Seb9Hj9h6dBr9Ug2eRbKcLLuk80WfDn9UfkCd+9rXpikVrJKFyQ5hnU5HKlhiqgYgLuLUhPDJJ6OrQG2PoioDMBM24MasrMTDSgxWKnXjUxBldpOAfrOnHhlPyA+++uagMAnFKUBgADr9SckEzCIaaeAKCyRRlG3BclkZQaYlB5+tPDuOnl7+B0hffP5t14jzO9JB2TC/0P9StOT4BGYKMUXtlcLm+vGIZBTZ0fTw2gKDXUqyY24Omn9CDjD7BXmoHntAItx4I+H41KiE2+kHrNFEm9ig7VBa+A2lXJqodOKU4DgIFVarpbgJqd7PsR80J+GDcJA8wc74qwQpOCGmJQeemrE9h4qBGHepFTvWnx6lETCgadBsUZrFW3W2QVUgBwsmV4GYetDpd80fLx1KSQUhNL9FrObesEDn+s/FwXvDImk0YlxBxWhwvfHmelzrecNQIAcKgu8PnS6XJjTzULaqYVswXegCo15V8CEIHs8UBKYDXJG3X6yS0CHT2RqTUU1BCDht3pllNAVkd4K4imLs/Kp1DhvhpBAB66hM0cO9ncPfDtw/uRBin3btRp5MCNE2tdhZu6bNhXE3nPilin13Lug2uYQsPpLaihUQkxx5byFticbuSlmLBkKgsSTrZ0o8fu/5x5pKELPQ4Xkow6jMxiozUGVKlR+2nCQJ1+AhBx2wkKaohBo031oe0O8A8aiBaLZ+O9UJlUwCa7XzmjCBdMZBVSnVanLPMPB2pVM5+8m1XGWlfhW/61FUuf+Qon42gQqZoWqQokLVA3YZ56Si1mt/V7gz6frNSQUThm4GMO5o7JQnaSERmJBogicKTBv1qzSzIJTy1Klf2GxoFUaiINalo9bQCRdn6noIYYNNSReKBVh83pwttbKmWPiPxYr8Z7ofLj+aPw9DWn4OFLJ8Ok1yJX8pjE+vTuj/bW4voXv8XJZkvAyidAUWoau4Z+V+Eeuwt7qtogisBRaahjvBFUqeluAY6tZ9/P/xW77UWpySClZshidbiws7INh+s7Udveg/oOK442dGLjwQYAwNyx2RAEAeNykwEws7A/dnGTsOSnAVRKTZiKeNi0VQAtxwFBC5TOCflhTpdbPm8VSOetSHvVUPUTMWioP7Q9Af7Z1uypxYp3duPyGYX4w9XT5O08/eThqXG7gbW/BPQJwAUP+32+FJMel00vkn8uzUhEfYcNFS3dmF6S3offZnB5+etyfHuiBcve2in36ClI9R2AyAfFOVwiWrvtcpfZocjh+k7wuCtW0mXRhgf+fhvvHXgfcDuBvCnAxEuB1XcAXfVAVwOQlOP3+bJoVMKQ5c43d+DjffV+7xME4KzRrE3FuLxkbD7ejMMBgpqd3CQsVT4BKk+Ns5+VGj4aoXAmYEoJ+WG17Va43CIMOg3G5iWjpt0acVdhUmqIQUPdYClQ+qmu3Sbdeio1bf5O9vV7Wa+OTX8GOv2fHLwpyWTG4ZMxXgHFL/o7Ktrwjy+PA/Cv1PCuwurHDFXUM24ah/ix9he8tNXbGwVAST1NvgIwJgEZI9nPQdQaGpUwdNl2sg0AkGzSQacRoBGAVLMeRelm/PjsUbJ/cHweU2r8FVd0251yD5tpfpWa/g5qPmO3YaaeKqTKp6J0sypFSkoNEWOEkn7qlLr9enf97ehhBuNUs+pkz/+hAKBuN5B8fq/HUJoxPIIa9UWfXwi9K5842clGNFvsqO+wYkJ+6KupgUYtrzfEaQdkOf3k7R1zOYHyr9j3Ey5mt3lTWEl3/V5g9Ll+ny+DlJohSY/dJU9P/3LFAvm85m+A87i8wOmnfTUdcLlF5KYYPRY1ilLTj+knUYyoPw2glHMXpyfI0+jJU0PEHK0hpJ+6pOooHsRweJCTYlbF5fwfCmBj70OAKzUVMVzW3W13yu/TmaMy5e15ftJPgMosPMTVj/2k1AQu6bZ1AKK06k4rYbd5rJovmFLDR4p0Wp0DO7F5CNDe48Dznx+TjfQeDHL1Y5Vkkk026pBq1kMQBL8BDQCMkTw1jZ02H9+JYhJO89g+IEpNw37A0sjS/0Wzw3ooNwkXZ5jlRpMtEY5KoKCGGDQ8PDUBlRopqPFRatjPslLjtAMnNyk71O4O6RhKpRLvWFZqeAl3gkGLP14zTU5VjM5J8ru/XNY9hCugRFHEQVVQM9QDsP7A4VLmm/koNTbpvdGZAa30P5A3ld3WBa6ASjHpoZMqYvoyNDAWeXVzOR5fexAvfH5c2SiKwH9/AvxhAtBWOWjHxi/qRRkJAYMZTpJRh+IM3oTPU63h4xHUqSdggJQarpSXngnowqtK5eXcxekJsik+Uk8NpZ+IQaM1hJJuflLvtDohiiIEQYDD5ZYnF/O5T6jeCjhUgUntrpCOgaefGjpt6LG7YDZow/01Bh1+wc9JNiInxYT3bp+D6rYejJAGeHrDlZqdle34ZF8dtBoBc8dkw6AbOmuc2nYrOqyKOhePSg1vMyAIXmlWgDXdAzzNmLmSUtN0GHBYAb1v+lGjEZCRaEBDpw3NXXbkB1Dzhg2iCOx7F6jdjTN3bsdr+mYcqLsGwCR2/9d/Ana9zr7f8Sqw4NeDcpjKRT20v8e43BRUtvTgUF0HzpDUWavDJZd/z/AqehgQpYafc0vOCPuhilKTAG0fg+6hcxYj4o5Qqp+4p8blFuXAp1N1sZMnF3PX/ahz2G3bSaCnrddjSEvQy89R0RKbag2/4PPKprKsRMwZ7TvQk8PL2NcdqMePXt2GW/61Ffe/F7y/yUDDTcL8b9PYaRtWDRJDgftpUkx6+UQvY5WUGqMqqEkpAMwZgOgCGg8EfF6u5LVH2LE1FnC5RdZm/9Ba4D83A1//ETMsX2COdh9urXmABTPlXwHrH1QetPvtQUtDyZ4SaZHVG9wsvKW8Vd728b46tPc4UJBqwqkjMjz2N0kLln5VaiwsoEJy6F2EOWqlhhcyUPUTEXO0ehiFnX734V4RQElBdfQ4YIQdEwyN0GmljzCXPideCqRKHoNeenYAzIhXKldAxaavhptoc5L9G4O9uWBSHs4clYnJhSmYVpwGQQDe2lqJzw419OdhhgU3Qc4ZxYIzm9PtodzEA0ovJj9SPk8/GZOVbYIQkq/GrOeN2Ianp6bT6sCZj6/HDS99Cxz4HwDAXXwGHnJej9eckoH60weAVy9nvqRJlzEfSOsJoHr7oBwzVypKQgxqFk3OAwCs3Vsrn7fe3srSZ1fOKvYJgo3S37xflZqeFnabkBl8P++HqUzSnp4aCmqIGENd0h1YqVEFNT2Kv+Z3+n9greZOYPNzgK2LpZ8AYOTZQL7kLQg5BcXSNLGq1DR4KTW9kZtiwus/PB0f/Gwu3vvpHNx4ZhkA4N539/h4lwYLbhKeXpLmodbEE0HLuf2ln4CQfDUm6QIX6H+uL2yvaB30AbF7qztQ32HD5qONcEtzsU5OXYaXnIvxG+ct+EfijwAIgMsG5EwELnkOGL+EPXjP24NyzLJSkRFa+mlyYSrmj8uGWwT++tkxVDR34+ujzRAE4KqZRT77mwZiTEI3m08VblDjbZLmnppOqxP2CPrqUFBDDBrqSDyQp8avUtNtx7maHWzjx78GPr6XNSFLKwXSy4D8U9h9daGZhWO9Vw03CvMJ3OGyYuF4lGYmoLbdikc+2B/NQ4sYbhIen58iB2vxVtYdtJuwVZqHZfQKarivpn5fwOflvrFA5vxIaeiw4sq/bsKNL38X1ecNF94dfLpwBJqeZsCUiu0YJ9+/SncRcN1bwCnXAt97HTAkAFOuYnfufZeVyw8goih6lDSHyh0LRgMA3tlehT9vOAKANejzl8LiYxL6Vanp5kpNRvD9vFD7aQRBQIpZDy40tUWQgqKghhgUeuwuj5WivxOsKIqypwZQKp5c9QeQJvBUkQhsf4V9O/JsdpsXrlIjBTUxqtQ0StJtdoTdgc0GLVZeeQoEAXh7axW2V7T2/qB+xOpw4UQT+/tOyE9WRjvEuFLz6f56PPS//fLv1hu8j5PfCd02P54aAMgey26bDgd83v5KP9W0W+EWgeNNFjhcAzBjKADl0vt7nlZKJY25AIcalYDYYncCYxcClz0PZLCp1xh1DvMjWRo8W0MMAO09Dnmwb1EYQc2ssgycNiIDDpeI/2yrAgBcM7vY775cqek3T43TBtilUSbhBjVeKpVWI8if+UiGWlJQY2lilQLEgOJtAvMnhducbjhcinGPKzUJdVsAAAdNpwDjlyoPGCEFNVypaToM2HsPVOReNbHqqZFKs3NSQvPU+OPUERlYMI611t9X09HL3v0LH4+QmWhAdpIR2ZJXKNaDmoc/2I+Xvj6B8/7wOe5etavXNA2vfsrwN8wyUPopcwy7tTQoao4X5n5KP3WpUsVNg9ixmCs152mkoGbsIo9GdX4VKq2eeWsAYM+q/j5ED/hFPSvJGHb15R3njJa/T0/Q43xpSK83/a7UcJVG0ALG1LAeWiedv9SVeOlSyjUSX018BzWNh4A/TARW/xQA4q66YjDx/rD6O9F0ehlDuacmrYn5Z04mTQeu+Acw6lwgfQQw+jy2Y3IekJjNTIBBZHgO71VT1doD5yCuMCOlUVXS3Re40tMeYdVBtOCVTxPyUyAIwrBQapwuN6rb2MXL5WYr6wVPfYZ7/rNbTj14E7DxHuC/+glgQU4SM5Gi6ajf5zXJ6afofta7bIqqylOikVDZ0h1xN1kAKG/qRolQjzGaajhELZwjz/WYk2QJUJSAqVez2wP/Axx+GvT1E+rGc+Fy1ugseXDlpdML5eDFm35XamQ/TQagCS+s4J8VtSeQm+NbI2jAF99BzaE1zCx2bAN2VLRi5iPr8Md1gWVbInqEotSo/TSAkn7KadsJAKhPmw7ozcD/vQP8fAdgTmM7CoLKMNl7CiovxQSDVgOnW4y5FJTD5ZZb3vc1qEkdIqW+B2rZBYiXrSqemtgNauSBfVoN3r39TJw9Nhsut4i3tlZiwZOf4dZ/bcGf1h3B+gP1spnT73wzjr/qJ06WpNYESEH1l1KjXoRE+rdqsdhx7h8+x7V//yaix7vdIsqbLbJK8517PLbUuWU1AACsDrf/CfXFpwHJBSyNUhHZ60dCJH4ajiAI+OM10/CT+aOw7LyxAffrf6VGCmrM4aWeAHX1pnL+Sqf0U4RUfMtue1rw3b6jaLHY8cd1R/DC58cG97jigBavC7E/o3CndxdhqwNor0KqrRZOUYP2TCnNJAjsSw1PQYXQWVirETCzlDWrenXzyXB+jUGnWZpWrtMI/i9+YcAbvA12UHO0geXmx/KgJin2lZqqVrbyL0w3Y0ZJOv5186l45ydn4KzRWXC6Raw70ICn1x3GLf/ail+uYp9ZpaQ7jPQToAQ1zUf8Hkt/eWrUi5BITd3lzRbYnW4caejyH3j0Ql2HFTanG+drtwEA1rtn4K0tFQAg9z8BAgR0gqDMLBpAX01flBoAGJGViHsWjfdt0Kii35WaCMu5AZXSrEqfK0oNBTWhI4pA5bfyj0LTIfn7x9YexNtbBq9ldn/x7fFmPPLB/qhXPUQC/7AWSh00/So1/tJP0gpqv1gKc1KQ3C0v6678LqSGWj+VKgne+K5iSI8P8IZfPLKSjNB4N2gLE35S5F6OwYKf5LmBm1d1xXL1Ey9bLVJ1jJ1ZmoF/33oa/nfHWbh/6UQsmcqaln26vx5Wh0v+O4SVfgIUX01TgKCmn6qf1P+vkaafWqQg3eUWI2ovUN5kQTK6MVvDzufr3DOwZm8dAGBqUaq89gmYguLFBurhuP0M99SE2qMmEgZMqQnTJAx4dkTnZPShV038BjVNR5ToEoCpnakzI6XW8r96dzeWvbkD6w/UR1QrP6CIItByAjj8CdAVuIHaQx/sxz++OoGNQ6DJWot0wi5IYyd5u9PNOoCq8G621mF1yEHNVvc4pARZmaBsLqAzsc6qxzf2ejxzRmdiZmk6bE43/hpDSl1fy7nVDAWlxuUWUSN5T4qkk3z2MPDUcKXGX3XLlKJU3HLWCPzl2unITzWhx+HC5uPNcoo25OZ7nCxeAeU/qOmvPjWeSk2EQY0q3RDJJPHy5m6MFyqggwvWhAJUiLny+XtcXoqSegsU0PFig5qdivm1n5GVmgjST6FiVCk1/eIdjbCc2+50+6j2AAU1kVHpmTNNtZwAAKxYNA7XnloMtwi8t7MGt/xrK858fENAM9+g4nYD794GPF4C/Hka8PpVrCW4Hyw2p2zAHApKBFdqitKUlav3SdbHU6MKara4xwWVW5GYBcy8kX3/xVO9Ho8gCLjzXLbCff3b2FFr/K1yImUoBDX1HVY4XCJ0GgF5khzNOyW3djuG/gIjAEpQEzjFIAgCzhnPKtA+3V+Ptp5gzfekoMbkR63MkipiWo4Bbt+Ld795atRBTYT/P+qLWCQXtPJmC0ZqagEA2pxxHveNy0tCgoE1crTYAvzuKflA1jgAIhuj0M+43aL82Qh1REIkcKVGFOFRURo1uiNLP/EqOe/0eXofhlrGb1DDjWCJ7CSSbWN51+xkE3532RS885MzcdOcMqQl6NHUZRtSLeRlqrYAu99kJziNTtnm50S2q6oNXAiJZAUUbfiKLC/VJEvC3V6SMPfUJEhyudPSDtSzTqlb3WOVYZaBOPPngEYPnPzKc4J3AOaOycKMkjTYnG688MXxXvcfCnjPfeoLQyGo4Sf4gjSz3Oo9zaxMlh7MUuG+4C/95I9zJ7Dz0Ye7a+WsaZo5zPRTajFTKV12NgPNC7NBWrX3Y0l3pEpNax+DmhNNFowUagAA+txxHkNdx+WmyOeSHkeQBnvcVzMAKaiGThvsTje0GgH5qZG3ZOgN7qkB+slXE2E3YXU3dHX6nJSaSOBBzbRrAQDFLuahyUk2QhCYcfT/XTQJ159eCoC13h5yHFrDbidcDNxbzU5kTivQWu6z646KNvn7pq7BD2paVXNtZOOiV4kpP0kWSmpOSfdeACIqkYdGpCPF3MuQ+dRCYPr32fdfPNnrMQmCgDulCoJ/birHS1+dGPJl/txnkh3i3KdgDIVBh8pgP+Xir9EIMV8BFSz9pObMUVkw6TXy3yDZqPOdni6KilHYX/pJowUyRrHv/ZR195qCiZBoGIX7rNQ0WTBSYEoNMkdjulTurNUIGJmdKAc1gTqYAxhQszBPPeWnmpQ5dv2AQauRF4/94quJNKjp8K18AiDPfyKjcKh0NTJpFgCm/R8AoBCNMMLus+KdVMDk3T3V/htZDSqH1rLbiZcAepNS9dB40GfXHaousS2Wwb8wqIf1yScar9VTp80JAW5cYvgOf9T/Bb+x/QEAsMXNAo9elRoAmLOMNYQ6th7Y+Bjw8W+AD+4C2qv97j5vTBaumVUMl1vEQx/sx92rdg/pwX8RpZ8szcDbNwBH13ls5kpNt901aGke+eKf5nnxj2VfjdOllBQX96LUmPRaeYgnAKT5q3xydLNJ3ID/6idASUH5qYAyDkjzPbuPRy4U+hLUuKWWDHJQkzUG00vSAABlmQkw6bXyuSZg+gkAyuYAggZoPgq09W/BSF/KucNBEAQYdf2j0AGIuKRbUWo8F2UZVNIdJrzqKXs8kDUGLmMaNIKIyaZG2UTHmVLEgprD9Z39OwwsXJqPAU2HWNqJN53LnsBuGw547CqKIrarlJrmoaDUqHpw8Pfce/XUaXXiUs3XuKPpEVyq3YQ0dEE0Z+ANx3wASl+VoGSMUJpqff44sPkvwNYXga//5Hd3QRDw+BVTcP/SidAIbK7Kve/2Pu17sIgoqNn2ErB/NfDJAx6bk1VB4mCpNYHKW2O5AZ/co0anQVYIoyzOkVJQQKC5T5JqLGjZdGl/yGZh3141A+GpcbnFiJQW9UUs3MfXdljhdtpRIkhWgcwxuPiUQpwzPgc/O4ct+LinJmj6yZQKFM5k3/ezWhPuIMu+IFdA9ceCJcKSbn7+yk3xVmrYucjqcIetKMZpUCOlnkpOBwQB3SkjAQBTTb6+mYJUE9IT9HC6RRxSdaUcdA5/xG5L5yhN53LGs1svpeZkc7fHCWKwPTWiKPpVaqw+QY0D87Wsed4613RcZXsAFTfvxBZxPAQBSDL0kn7inHM/MO5CpmiNk6bxBvHYCIKAW84agWeunQEA+PJIUzi/3oDSFImn5sQX7LZhH9BWIW/WagSkSBOxByuoUbwn/pWaWCzr5oFaUZo5pLJ7bhYGepv7lOzbn4kjl3UHTj95/7/1FXVHYSCyv1VfPDXlTRaUCA3QCy4W7KUUIDVBj5dunI1LpxcCQGhKDaDy1fRvUHOkgV1T+qTUiGJIbSvkXjX9otREVv3UKDfe81Rqkow66LXssx2uWhOfQQ330xSfDgBoSSgDAEzQ1fnsKggCJhcytWZI+Wp46mnchco2rtR4BTU7Klnqic/TGGyzZZfNKTvw0xMUT423UtNldeAMDZsa/bJ7KbaI41HZpngNQu7LkloIXPsGcPUrwFKWwkL9XqCnLejDZpexhnwtFltEUnp/I4qi38ZVQXFYWe8ezuGPPe4e7K7CgVausTz/Sd14LxTyU82YmM/SSn7LublSEyj1BARNP/E+NdYor9h5+ol7gCLxPzX3Iag54eGnGeU34EsItUcPL+0++mmv54lIsdic2HCQLaTPGBV+0zqZLf8AHsoAyr8Oulu/KTV9GGYZqCWFICjVUOH6auIvqHFYWQ8CACg5DQBQpy8BAIwU/PsseFAzZHw13S2K0jBukbKdKzVNRzwqoLafbAMAnDOeDTvrtDoHNZXG53mY9VqYDVqlGZjXCiLFUo4coQ0ujQHHjCxg4yv5oD1qgpGcB2SMBCB6Xtz9kJFogCAAbjGy0sL+pq3bAbs0qyrkCd1VW5iZnMMVPwnuq+kYhKBG7T0JrNTEblATzgTmi04pAACMzknyvVNWaoI0n+RKTVe9z2DL/jAKi6IoG4VHSLPUwi3rdrjcHqMWIlFqeOWT/Pt7kWBkSmRQozDAVPyssUBPK7Dut2EdR6h8vK8O3XYXyjIT5I7mEbHlRTbn7sD/gu7GlRpbtJWaPgyzrPczIoHDA3r+/xMq8RfUNB4A3A6W+0tnY+crNEyaLHT6N4VNlszC+2qGSFBzdB0zCuZMBNLLlO1pZYDO7FMBtV0yCc8fly2XxkYyKCxatHg1FQt0kh1j2QEA6MyeAZOZXRD4Bzxoj5reKDmT3VYEL/PWaTWyp2EoKgT8Ap+eoPetkAlE+Zfsls/GOvEFYOuS75a7CvcMfBCn9p54B2mx7KkJtZxbzY/mjcTrt56GW84a4XtnsMZ7nCCDLdXN96JV3WdzumX1dWQ2D2rC+1t5r8jDDmqaLRghSGp7VoCgRlaFg3hqADa1e+kf2ffbXu6XWVDvbmeL6MtnFEEIlEbsjfZqdk0DgIb9QXftN6UmCsMsvdNPADC7jKk+f/38WFgjM+IvqKmX/vB5U2V58oiLrYqybJWsoZ0XUySl5mBt59Bo/sVLucct9tyu0QDZkkFQMgt32504KHmBZpamy4HEYKag+MmLm8EU855nUDPRzvw0PYVnysqMrNSEUvkUiNIz2O3Jzb3uyo2dg52y80evPWqqtwF/PQt4cSGTiAHFTzP7FhYQu+weZkjeE6V9EEYlBPOexHL1UyiN97zRagScOTrLp3ABQGjpJyDgDCiujALRu8Cpy7l5b5hwVTVv70T4QU233Hivz0oNwKqgpl/Pvv/fnYAzeoF+bXsPvj7GvHqXSX6fiDi2Xvneq0DEm37z1ERYzu1yi/J51V9H9J+dOxqJBi12VbZh7d7akJ83/oKahn3sNm+yvOmQLQM2UQe92wp0VPk8pDjDjBSTDnaXWzZ2DRqiCByT2v6PXeR7v+yrYR/w3VXtcLlF5KWYUJBmRqZ0kR5MszA/WfGcqdlf7whRxDQ3a7TnLp0rBzGV0gWi1x41wSiRgpqa7SwdqaazHnhhHvCviwG3G1nJQ1mp8W+yg9sNfP1n4MULgPo9zBi/5UXAbgGqtrJ9RpytfH64PwtKWq+9p5eVbD8gX/z9dFZVD7Uc6r2DvKmOIP0UFLlHTYhBjZfHzqRS9aKVguJ+miSjDnlSE7lwjcL8vMA7Sfc4XCEfn8stoqK5W0k/cU+RF7JRuDelhnP+Q0BCFnsPNz8T2mNC4L0dNRBF4NQRGX3rJHxUFdRYGgBL4KKGfldqwiznbrbY4BaZtpDpxzuWk2zCT+azfktPf+p/4rw/4i+oqZeiWS6/A6jvcqJc5FKt75vnaRYe5BSUpRGwtgEQPH4HGe6raWAnMp56mlGaBkD58DQPplITKP2kWkG46/cjEx3oFo3Ql86Sg5ioKDUZI4GkXKZSVG9Ttve0Av++HKjdxdSLmh3yxbSpywa4BnfQozd+y7lrdgKvXAx8ej/gdrIUJQB8+SRw5FOWek0tZioND2qOfCIrlIOZfqpqCZym4UqN3eVmg01jBIfLjdr2KJfthpJ+AoCC6ez2xJcem3UNe3G17isAYtTKurlSk2TUyZ/HsJUaKagpzjDDIDWiC7XypbKlGyZXJ7IF6b3JDB7UhBzMJWQAFzzMvt/6z5CqjHpDFEW8u50tnq+Y0QeVxuVU5trxjvJB1Jp+U2p6Iqt84qmnzERjwMaDt84diYJUE+rCSGXGX1DD8455U+RNjZ02HBNZCgpbX2bzk16+EGhUApwhUwHFg670UtZwz5tsz7Lub46zD9yMEmZEy0yKvP10X9ha3oJr//YNPtlX56PUKCca5WLlOMZSIlvdY5GSmCgHMfXShztiozDAlgZcreGGa7sFeP0aeQwDAODQh3L6yVTzLfC7AuDzJyJ/3SjDBz/mpJjYINN3bgX+djbzzehMwNKngdu+ZFJ8dzPw4V3sgSPmsfegdA5gSGJm0tqdAAa3q3BNS+DyVpNeKwe/gxFwRUpduxVuETD68QlFTKjpp9Hns9vqbcoK3u0G3rwOT+iew0LNlqgFNdzgm2TSyZVqkXpqMhINSpv8EHtqHWvsUiqfkvMDBnw81R1K+snudMPhcrNWEFoD0F7BGvL1kT3V7TjS0AWjToPFU/Ijf6LqbcwEbkoDRp3DtgUJavpPqYmsR01jCD22THotViwaH9bzxl9Q47AAWqOcc3W43GjptitBzcEPgL3vACe/ZgYxiSFTAcWDGt5cyxse1DQdgdVmw3cnmDQ4d0w2ABYVAwM/KuH1byuw+XgzfvTqNqzaxlYp/MTlr/meKK0uvxMnwqjT+AQxfTIKA+yCDjCzsKWJBTSV37LGW3OWsfsOrpEVghlVrzJlZ89/+va6UeRgLQsCxuUlsUB8zyp2x5SrgZ9+B8y6GdDqgHPvZ9u5TDxiHrvVGZST4ZYXAQxs9VN7twO/XLULXx9tAva9hycPnoulms0BvSc84GobBL9PpHCfUGG6OXIzqDehpp9S8qXFm6h0j67YDLSzgojrtBuil37yo9SEmypsloMao9wmvznE7udHG7pUlU/+VRoAqjEJwdU+i82JOb/fgMuf24QemJRFkDrdo6bhALDjtZCUnI0HGwEA503I7ZvizP00oxYAuZKdIohZ2NjvnppwuwlL6XM/fho1F59SgMkFvXzWVcRfUAMAORPYyR5MsRBF4H33HIg5E1nTpfFL2X689BuQ39QDtR1wuqIU6VZ8A2z6C7DhEeDDXwBvXMuMnX+YFPji2SSZ/gIFNWmlrPGUy4Z9+3bD6nAjJ9mIsbmsNJQrNQOdfjreZJG/5xE6P3EleJd0u93QVzIFZbd+KgRBaQrH8f45bLhZuOIb4Pm5TN3QJwLf/w8w505Wnth4AKVCPbLRigldUvVD0yGWphpk3G4R+6Wp6zO0J9jxa3TArRuAK/7OlDzOhIuVDqkAUDZX+f602wAIwM5/A3v+o6SfBiBwePGr41i1rQp3vrkTzp1vAADO1W4P6DHgjejaBnE2Vbj4LecWReDTB4CP7o0snWGTFla9KTUAMGYhu+X9iHjgC2CuZg/cLb4DLyOBN95LNuk8UoXhKH6KUqOX0+ShtlI42tClmIQDVD4BCG32E1jPm8ZOG/ZUt+PxtQeA0eeyO475CWpsXcArlwCrb/dpkeCPY42s2pB3q48YHqiOOldJM8eQUqNUPgUPajQaAU9edUrIzxufQY1X6gkA2hNHQrh9M3DDamDBb9iddbtlr0FZZiKSjDrYnG4cbezyecqwaa8G/rkE+OQ3wBcrWQOlQ2uYsbOjCtj6kv/HyUpNgH9cjUYOeMoPMFPo3DHZ8ipR9tQMcPrphBTULD9vLJKkCoQS6eJl9s5zV2+D1taGLtGECpM058lLmelT+glgJwFjKpuj01nDlLsfrgeKT2UrjlJW9j2m7Utcqf0SWqhOBGofziBR2dqNLpsTBp0GxQeZyoLJVwJFM313FgRmeOQ+rFRVHr/sLGDeL9n3/7sTuQ6movV3+sntFvGOVNLa1GWF/QQbXVIq1AdWauSAK3bST359QvveZWM6vnlOqUYLh2ATur0ZKwU1x9YDjh42HgNAi8BGw6QdejP81/eD2ihs0mvl4Lg+jBRUixRIpycYFKXGn6LcfIydM62KFcAj/RSg8gkIPf2kDur/tfkktupYd3GUf6VUEnK+/hNL4QJKZWoQ+LlwpGqCeNh0twDV29n3o89lC3WABTUBAuWhVv2keAJ7bxzqr3ggEBTU+CuLzRrL+r3Yu+QcqkYjYKKk1qh9NTanC/e9twcf7fXtRhyUgx8wI2dqCTD7VmDu3cCSp5gPAgBqd/stL+81/QTIH3BLFfOHzB2jDMiTq58GUKlptdjli+SP5o3Ex8vn4YXrZ2LuaHZcPkbhL1YCAD5xz0KCiV0MvGXaPsm2AJtkzFdfky4HfrRROTEAwHg2TiGvdj2u1kpmPFMau+UVRINFdwv2S2nQuVnd0BxgFyqceUfgx5SdBdz2BfD9Vb73zf8VU2/sXZj41c9ghD06QU1recBxFN+eaEF1Ww+zNwkNSHAy9atMqPdbCQEMjSni4eJTzm3rAj6+T9lh64vhP2mo6SeAKXQJmcx78dnjzNSZmIOXk38EAMg5uooZTvtIpyr9BEBlFg69AooP2s1MMgRXar54kqnbH90LgBlvWfopHKUm+O/s/bq3f2qFOzGHLYLUPWvaq4FNqqqoI58GVd9EUcRxaVHM+/mEja0T2PgoAJEtzlIK2O8saJmK11Hj92H936fGf1DT3u1AvZ9GjKGmn8Il7oMa/sZ6BDVanbJP7S558xQ/FVAf7a3Dv7+pwG/+uyesBkFy98fTbmPBzLn3s+Bm+g1SQNWpTBLn2LuVqbHBghrJV5PexQKyOaPVQU3/KjVWhwu/emc3Vu9UujPz1FNBqglmgxaFaWYsnJQn9yLxKOmu/A448jHcghbPOC9DkpRm8i7hDmmYZW9c/GfghxuBK1/yNRZK4ycSa7/FCE09OkUzXNxr00sn4n7l5GbgiZGY9smVKBbqcZN2LesmOnK+x+faL/lTWUdlbzRa4Ip/AInZMLccwMXaTdFJ8bx+DTPc1/vm+d+Rqj+umlmEpelKG4UMoROC1b9vjQc1g9k4MlwO1bMARF6Vf7GSKYOJzOOGgx8CnWEuiGwhGoUBKXiXBt7yi++ky7AzaR6axBSYrA2s+q2PdKmMwoByoQrHLNxiUZQa2Sjs7zzVLZmed74G1OxEU5cdebYTGCWXc/c9/cTVwLPHZmNUdiIauuz4TjuN3alOQW18FHD2AEWzWdq/sxaoCzwAt6HTBovdBa1GQElGmEGNKALb/gn8eTpT9gFg2vfZrc6oeIkCpKD6T6mR0k8BSrqv/fs3OP8Pn/v0+opoGG8IxGdQkztJ/lZWarwrE/KlHJ5UFQIAkwu5UqOcdLefZCvMZosdu6raQnt99ZiDCUs979PqlB46qoDK6nCh7sQ+ACJgTg8u9UkXt0lCOSbmp3gEbFmJXKnpn6BmzZ5avLmlEg9/cEA2CXK5dUSAlYk80NLhYiswAOWFF+OEmC97Z6Ku1AAskCmc4X8oYHqpYr4D8D/X6WjPl8zF1Vv9q2gDQc0OACLyu/ZhjeHXOK1NCo7P/Fnfnjc5D5h5E3sqzT7Yne6+nfzaq6QKPJEZsFV0251Yu4etqq+aVYzrCr0u6q0n/D5lqpl7amIj/WR3unFYCmomFaQyP9zmZ9mdFz/DZs+5ncD2V4I/0eGPgX9fAXRISoScfuqlpJsz5gJ2K0p/zylXwmAw4T8uyTC+7Z+hPU8QuFE4WVJqcnkFVBhl3dxTk6k2Cvs7T9m5P08EPv41jlfX4zn9n9ggy9HnM19hANTpp2Am5lYp/ZSfasLT10yDViPg9SYpWDq6gd3W7AR2vg4AsJ73KHqKJa/akY8RCO6nKU43h94FnFP+JWsCaGlkbSmufhU446fK/XIKyr9ZuN+UmiAl3TanC/trO9BhdeLLI40e9/GANzuE9FM4xF1QY0sqZBUuEgG7shZMY7fcLCyKOH/Xcqw3/AK1NZXygMNtFYppdP0B3ynffjm0lp1gcqd4jjng5PPX3iFvuv+9vXj01ffZD5ljAk/nVT1+pKYO54709CdwpabH4eq9VXgE8AFtTV02OZ9eLgU1ZZn+gxqznp1oxnTvYP1hNHp8W3wrAEXO9vXU9NEoHAqqYaFvuRag1jSSqWjW9qiUdkaEJPU6oEOy0AO9y8ok6FHn9v25y84CAJyuOQBA7FuaRy3R1+32uOujvXWw2F0ozUzArNJ0FHay+92i9JluOe73KflA1sHodhwJh+s74XCJSDXrUeQ4Aay6ifUJGnMB6xE0+xa247Z/Bk4BdbcA/72NmUJ3vsa2hZN+AliFmyCd6tNKgKLZMBm0eMu1gG07+qmy2o4Qb6UmOyW89JMoikqrh96Mwnx4IgCc/BpjPrwSozU1aNVmApc9H/TcmGBkF3aXW5TnpvmDv25aggFTi9Jw27yR+MotKaH1e5jq9a+LAIjA5Cux7EsdHj1czO4/HFj5kv002X5mevVG4yF2WzYXuP1bYOLFnr9rL2bhflFqHFbVMEvfhba6yvarI83y9+phvLmUfuobXanjPH5u7AoggfHAonYXW5XX7EBS+ScYpanFCvwTJ5q60G134kCt0mF4/cEQgxqeevJWaTg8oFIpNTsq2zBC5PJqkNQTADEhA7VgKafz0+s97kswaGGUVgjRVmscLjc+P6xE47z8XVZqAhjjePrpBz3/Zhtm3ohaTQ4AIFlSZPpFqemNyZcDGj0OaMdhlzgKTd2i0tCsapBSUFJQ87xzKf7ovBzutDLggkeCB7mhUjQb0BqQL7SgVKiPYlDjKcfz1NPl04sg2C0Q6lmX7+PJkhkzQFAjl3QPgqfmH18ex63/2gKLLfSFwP6aDujhxP2JqyG8cDa7GJrSgEWPs7/XhIuZZN9RHXh1//kTSrVdw37WANIpDfgLJf0EsBV0MRvei8lXAIIAs16LE2I+2swlLH3JTacRonhq2N8oJwSl5pXN5bjhpe/QaXXAYnfJQUZGokHuYeU3Tc6VGqmKL6PzEFyigA/GPgokZvnuryJBNXai2xb44t4um5bZ7/Pzc8cgI6cAe9xlbIdP7mNpwKLZcF/wO3x5pBHrnJK6X7UFsEgXcFunx3iF4419MAm3VbDbvCmsHYM3XKlp9B/UcKXG6oiiUsNVGkHrIRZw1J3Yvz7aJKtj7T2qYbyUfuobLYmePQwUpcZLAssezxqY2TvZSZavkgBcot2Ehu3/w65KNoIgPUEPjcDKvXlDtIDYuoBjknw54SL/+3gHVNJxjtL0njMGgCMNXdjlYoPwJsDzAiEIQr/NM9pa3uoxZZcHNcfl1Ung9FMOWjFVPMj+Oeb+Ap1WdlLx56nRaQQ5ZdWv5EwA7vgOf8x7DIDAPivFs9l9VVv6//X9IZ1EmsRUrE77ATTLdimG575iSAAKZwFgak2fyrrVQU39PnlqfEOHFZuOsRP+5TMK2agK0Q2kFGH0bGmWWUu536fk6aeBnpjudLnx9KeHse5AAz7Z75kqO97YFfD/aF9NO36lewNXdv2bKTTjLgRu3wxkstbv0JuA6f/Hvl/3W2DXmx7DRdF4GNjyd+Xnur0eFT8whJh+AlggdeqPWLsCKOb82iRpdV/Tt6DGx1PDe9UE8dT8a1M5vjjciE/21ctN9kx6DRIMOllR9h5yCUAJahb8Rh7a+ZTzahhGzun1OHVajdytuDuIYqEoNXrpuLR44sqp+Nw9DQDg1uiB8x4Ebv4YVY5kWOwu1CET5bqRkPsC7X8feGo88I9zZfMwNwkHSsUHpU0qv08r8X+/rNQc9Jsel6d0O6Oo1HSrUk9+FlbqoKauw4pjUlDHg920BL0cbEWLuAtq6s2jPH5uCJR+0uoUT0XlN3J/h4pENppg0rbfYs9xtuI8c3SW3LG3V7Xm6DrAZWMTwvmH0JvscaxBoK0DaD0Bm9OF9h6HbIRrNAXOGQOssmSvtKLQ1+/2ub+/ugpvPMR+d64E7a1uhyiKIaSftCgRJEUptQhIyZdPksnSSdKs18oTxlPM+ug1MuuNjJFITGGyalOXjakZwOBVQEknkVYxGRPzQ29IFTJyCmp/5EpNT5vSmVmjZxUjzcz0Xt7cDVEEyjITWD8abroung1kSBOpe1Fqop5+ajoCrBwDfPmU37sP1HbCIhlLPz+kKJHHGruw6E9f4saX/at2e6vbsVQrDU298Enge6+zShU1s29hrQWaDrM005NjWGfoI5+ydg9upxxoovkI81MArKeSNowUbME04MKVzI8H5QJXZZZW931Uarw9NTx9FGzMATfrbj7eLO+XISk03Cjc1uOQU/0yPKhJygF+8D4e0N2Fv7ouwuic0FI6PAXVEyT9zj01vDcSwLqyd834MZ5wXI2HC58HzloGaLQ4UKcEmmvt0uia9Q8Bb1/PUjN1u+WUkLzAy4og/cSVmkBBTcYIdt1w9gBt5T53y56aaCo1vVQ+ec/M+/ooM3kfqWfBXbRNwkAcBjWV+hEePweddMzTQF+sZD6K1GJsn/cPVLizkeqox4jdrPx6Zkk6zpnA0iUbDtT7Po+agx+w2wlLA6cMtHrFLFyzA42dNghwyyWLmzuCd24sb7Jgryj9nqoUFiczmAmvD6yXfvfvn8aCrj3V7ajvsKHHwdz+gZqqmQ1alAgsIBIlkx9XfPhJUhAE2VfT58Z7YZKlnhDNg5qG/Uxadjk8V9f9jRTUtCBZbjEQVVS+mvZIFZGqLQBEZmbk/0OSr4Z3Kk7lFwuueBWd2mtQk95fzfd2v82GAW56xu98ry3lit/kyyNNcpXj/3bVwO50Y291h0+g5XKLsNcdQK7QBrfWyKY9+zWklwE/+QqY/2v2fjm62QLqtStZVZJGB1z2AktTiW4l7Rlq6ikAXKkpN0rp+OptfZprJHcUlv43M0JYOMlBzbFmWZHhBmHek0gUvZQ5UVQ8HIYkWFJG4ZWuWRChwagQfSo8BWUJkn7i1U/pCZ5pnvnTxuI516X4qCFd3nZQZUFY55BSUHwwMm8Dcfgj2JwuVEp9i0ZFpNTwoCbAolajVVJQJzf73N0/Sk1oQY1WWpB+JaWg/vYFW+QsGJ8TvWOR6Leg5p///CcmT56MoqIizJ49G1999VXAfaurq3HNNdegrKwMhYWFWL58OWw2zwjvm2++wdy5c1FSUoIxY8bgb3/7W0THVSMqAYHF5pT/sfwGNTwN1FrObqddh/Gl+fi1k5lY53e+jyR0Y2ZpOs4dnwsA+PpYc2ADbneLMhF5fIDUk/dr1+5EY6cN+WhBgmCDXdRibVXw6LaypRt73dIFoumwz0U3g49KCLEFeSicbLbgWKMFOo2AH589EhqBfaC/Oc4+9CUZCdAHGFpm1mtRLLAVqCuVrULklZ/KOyNXQvW18V6YeAy1TM5jvYVEN/DW9cDK0cDjJR7dp/uDjQcbcKiuUz6JtIpJ/RPUFM2GQ9AjX2iBu/lY7/v7o0I6oZacoQxdlYLrDimtmGLSsQsUD2qKT2UXdQDoqlNVuCgoYxLs4bVP6I2TX7PbnlZmVPdCHdQ0W+xyJ+e1e5RUFN/GKW+2YJZrJwBAKJ3jf04bJ60EmH8P8LPtrCP0qbexydAAcMYdbOI0r9jkF6tQK58CYJLSt8d1o1jK19LAvD0R0qlqvgcoiktbt91XaZHgzTar23rkylGu0Oi0GlUJvyqocVrZ/x4AGBLlaqKsJIOHqhKMBGPvDfhavTw1nCmFqdAIQG27Ve69clBSarQaATvEMWg35jOF8uJnlBElhz9GZUs33CJ7j8L2kdi6lAAirTjwflJ/Lez17UjPx9FE1VMjT+hO93t3Yxd7j+ZJvdK+OdaM9QcasKuqHWa9Fj+cOzJ6xyLRL0HNq6++invvvRerVq1CVVUVVqxYgSVLluD4cd8VmN1ux/nnn4+ioiIcPXoU+/btw7Zt27B8+XJ5n4MHD+KCCy7AsmXLUFFRgdWrV+OBBx7A22+/HfaxqSf88igywaCV/xk94KtMzrTrMDo7CVs0p+CEOxd6uDBPfxATC1IwNjcJRelm2J1ufH1U+kN7r3y+fZ6tMnIns5N4MFTVV2o/zUkxD18ea2eD1gJQ2dqDJqTCas4FIHoOaQQ7AQDRVWp41dPssgzkpJhkKfj9Xey4yzIDd4Q067Uo1rCgxp7M/mFlT43q76IoNQMb1GQls/dL9k4USemA4xvZxHTRFVJ79Eg5XN+Jm/65BRf++Qu4pMGErWIyJvVH+smQgOoElhZNa/iWfYb3/VfuRBsS3E9TcjrrjQPIZmGu1KSY9UyR6W5mknneVHZi5CfHFt+ybt6l1i0qxtQ+47B6phL3vedxtyiK2FLOjLq8SuPzw4042tAl96ABfIOavdXtOEvD/u+EUfNDOxZBYB2hL3wC+MVBVuFy3m/ZfTwVXiG1ggi18ikAXKnpdOmAXCkN3ocUFB+TwP9feYDhFv03S3S63B7VR2ukEv8MVePFDH9mYXWwq0+Qg5pQVRqg9wZ8LrcoB9/egVKiUYexuSyg3FHRBgA4WMc+B4sm5cENDX6R+jTw8x3AjBuUMRVV36GiivUYG5mdGH76XJrXBVOaX0OuzJQr2e3xz4BOz6wBtwVEVanhnZST/Csu/Bp79thspJr16LQ58at32bng/04vkf2d0aRfdPwHH3wQd999NyZMYFLYVVddhVdeeQV/+ctf8Ic//MFj37fffhv19fV47LHHoNPpkJaWhqeffhpnnnkmHnroIWRlZeHJJ5/E2WefjSuuuAIAMHHiRPzyl7/E448/jquvvjqsY+tp6wQs7B+juaEFZrsVxclmeZsHCcWAy8A8MKVnAoZs6Kw9mJahx9d1kzBCV4crEvZAb2Xm4EVlSXitvhVf7yrH+e0fAp/+P2DhY8Dky1j66ovnALsIzLoD6O4OfqCpY9m+J3eipbQF4x0nAYioRh5c1i7sPFCF2SN8JT9RFNFU1wKz3Qln2iSgvQ449i2QqTRmy9U6YbZb0dnU5v/3joCvd52E2W7FBaVJgMWCGZkGVFZasWVvJcxuEWOTNAFfSwOgzFkHuET0aLKRYLHA0dEJs92KVLdNfly2wI47W3BG7bhDIVfj8ny/TrkFaK5hq2enDdjxKnB8CzC7f47pRHkDzHYrEtEDLdjJNkGfgGxN/7wP9eZTUNa6E4U1XwH/OQnskKrSTv8JcM79wSutnDbgxFbAKQKZp7Ag3i4CFTuBri50t3bAbLciEw7gu9fYfUVTAJuDfSWWAu0tQPUBINkzVWwCkC7aYXW40dHYhtQggXLInNwM9FhZybPoBna/Dyx4hKWAAZxsssDS0o5UnQY/mVWK3390CN/sqYC22wKzXSlXPnq8HpiunNiPHa3Cbc59gEsE8k6L7O+UWKycJ1JHs/eqXgr2hIQ+/e2TnDaY7Va4u7qAjClAxW7g6GagNHzTucPlhmDphhlAsssGWEQYAORonOi0OtHa0IoML79Lt9Xh8f5VVlphBjs38d8rX+dCrd2KjsY2IFdSulob2PugTwB6enCygv1vjE8JfH7xJl20w2y3wtbeCVh8g6EOix0mGzu2NNEOWDyDn1NzDDhZYcW+IzWYV2hGfW0zzCJw8/RsbNh2AltOaGBFKkwWC6DPANImsHT17g9hto/EuKT08P921YfY751eGPyxxlwgazozfm99Azj1h/JdZoeVvecWMXrnjbrj7LgMOVi/5TieXncYj18xVR4A3dHUBrPdikK9G/MLE/DJ/npYWqxI02nwo5l5oR9HOMcrRpmTJ0+KAMRDhw55bH/hhRfE8ePH++x/0003iddee63P9vz8fPHNN98URVEUR4wYIb7wwgse9x86dEgEINbV1fk9DqvVKra3t8tflZWVIgCxna096Yu+6Iu+6Iu+6CsGvtoBEYDY3t7eawwS9fRTTQ1LNxQUeLr8CwsLUV3tm7Otqanx2dd7f3/7FBayoXz+nhMAHnvsMaSmpspfxcVB8pAEQRAEQcQ8UU8/6fVMttVoPOOlQDlEvV7vs6/3/v726S0nee+99+Kuu+6Sf+7o6EBxcTEu/vXb+OzXrFPssxuO4i8bj+KqWYV46JJe5uao2F/bjiue24yX9b/H6dqDwAWPArNvht3pxsyHP8UPhffwc/17QNkc1leCNx9LLwNu+zL0Usz1DwHf/BUA0CMa0Jo1HfrLn8e8Z3YCAL66Z4E8oJKzdk8t7np7F6aXpOH1a8qAZ6Yzaf3uI6wPCVi+/6rnNyMnxYDPf3lOyL93IPZUt+Hq579BXqoRG+9mXUq77U7MfmQduEdw/S/ORkGa/+nLaD4OPD8H3aIRe6/biumlGZj6W9aVc9OvzpErIjYcqMfDH+7H45dPxWkjw5sI2xfcbhFTH/wELreIz345H7kpXqbPF85mhuyrXgHGnh/111/+5k58tK8Oz5zaivN23wXkTAJ+uC7qr8P5dF89Tv7nXpxjOoSRN76g+DkOfQy8eysrM772TWDk2Wy7vRvY/i/mR+Hdg899gKWrAGD1z5hxcd4K/LxqPnIOvYb79K8BhiTgR595Tg3f/TZrBV82B/i+l9nR0oQ7Xt+O9RVuPHn1KVgyJb9vv6jLwfqIOLqBW9ezmUKvX8N8PXfuArR6LP7TFyhv6sbz/zcDZ4/LwcMf7MPr3zJ/w5zRmfj7DbNwxmMb0N7jwDu3n4GJ+amoaetB/TPnY7rmGByLn4R+xvf7dpyc5+cqnaxPu03x20TAV0ca8cNXtmFcXjLe+/FpwJNjWSnwj74AsoP3wfLmYF0HLnt2EzKTDPjqHuV88pN/b8Nnhxrx4CWTcPUsz0UlP4dmJRtQlpGIrdK4mT9dMw0XTGa9Z57+9BD+9sUJ/N/pJfjNEsn3c2wj8OZ18v/AWb9fj+YuB969/UxMCNFj9pv/7sG726ux7LwxuO3sUT73bzxYj9tf24FJBSn4z0/O9Lnf5RZx6qPr0G134bQRGfj2RAtumlOKFYsmYNXWSjyweh+mFKbgtnmjcKypCyMyTDjv04UQupvxA/s9uOfHN2FifmpIxyrzn1uBQx+yRpu8E3Uw3riW+Wrm3g3M+wUA1iPq7JWfQSMAex9c2Pe2GC4n8PsyQHQBP9uBi/55BEcbLZhYkIx3fjIHnVYHTn2Uzcnadv95SDDo8OWRRiSbdJhW7N9YHJCODsCP+OGPqAc1RUVFAJi6Mnq00uiutrZWVle89+fqjhr1/v72qa1lxjJ/zwkARqMRRqOvCanBrQUSE+XvewwmJGemydtCYfzIBFw2dywcrecBFYeA6i+B+T+DAUBuQSYWtm0DNAJw6g3AyLNh+cs8JNob0XLmcmSkhPFhXng/MOYsLP+kFe/XpuHZhadi0Yg8ZOdVoKKlG0ctIjJzPY+73Cqgx2BCTl4GkD8KyMhjZq6ucrlxXEaOBj0GE6rtAsSEhD5/uFvQjR6DCcbUZPl9TEgECgqzcKShC0adBnn5mew98UdtA2AQUOnOgUVvhkVnRI+BBQ5JmamAVDV1zqyROGdW9N3yvaEBkJCegoZOGxrdOuR6f1bKZgIdR4D2g0DipVF//XIb0GMwITvZBRgEID07rM9ruCRnpeIRzQ14PSURG0aeptwx43Kg/ltmeP/yYWCiNFPo3euVIX9GHWsGePqNyjGWzQAOvwM078JpHa24LOEd5stZ8hBQ4NUdu3Ai+x0tlZ6/Y0ct8M8F+J3NjdMMT6FZ1PX9PajaCgg9QGoG+xuKbiAti5mXG7ejIedMHOhwQzCaMG1CEWDW44wppXhxBzO1nzNzJISkJIwozcHm483Y1+bCxNGJOHikBvOM5YAgQD9lcfT+ViWnAJ1SRVpqZp+e15hmQ4/BhHaNAUhJBUqnsX5c7QeBsmlhPVenlj2XPiXR45iSMlLRY+hEo9v3b9WjZ4/RJCVg+sRCfFnLfIkp2WnyvsmZaegxmFDvUj1e52afj+QUuM0JqHFo4TZokZmTDiSGNkNIn5KMHoMJHVqD3/ewGQb0GEwwp6f4vV8LYPSIXHx7ogWfVXcDBhNGleUBiYk4dXIxetYew3eNdnz3jtLZ98XU6TjXsB5zNPtQVpIDGIJceu3dbFCmoxu48ClAowGs1ez3zh8T2t999nVA1efA0dXAIuaDM2oM8nnVYUoIf/aUN60nAb2bGf1zR6DaeRw9BhOOdIlAYiIau7vQYzAhyahDQjq77s2dFuFn1hW6uTnq6afc3FxMmzYNa9as8dj+6aefYvHixT77L1q0CJ988gmcTsWMdfDgQTQ0NODcc8+V9/H3fNOmTUNubm5Yx9dldcklhk2Sq56XOIeKRiPgd5dNwdwLr2Ubyr9kBkkAZ6c1YpymCi5BD4xbDDE5H9eKD+PH9mX4R8fpYb0OjEnAhIuwxVoEF7Ty5NsSqd9LRYuv2biytdtjH3+DOXmFgdMtelSDRUqgng58qnlZZqI8kdv/E7BOmZViNnocLrk81KTXBCwDH2i4S7+xy4YOqwP/3VEll537zAmLMrVtzLSYpZHMcn4Gx0UTXmXkt/ne2fcwJaNhP1NnPrmPBTT6BNZg7u7DwPdXebar5xVQRz7B9zv+gQTBhua8ucDMm32fn5d1t1fJ/1MAgLW/BLqbkeJqxRzNvr51O+bwUu6SM9mFQ6tTunxvfxXbpKqncbnJ8ntyxqhMJBt1MOu1uGASO/fw0npeAdWyay10ghsNhuLg5bfhohrEG63qJ15WjUJpRMWRT1lfrr/NBz5fyRwNvdDlVc7NCTZpm5dTm/VanKFSXTPV1U/+Hs+rnwyJaOtxyEpweqLnuScYvU3qDnQ+UzOtJM3j5/H5rCKqJCMBc8dkwaDVYEJ+Ci6ckocEgxb/6WRq50L9Dnmopl8aDgJ/XwBs/guw9SWl5UFvjfe8Gb+E/U82H2UNX6FUPwFRqoDiFVmpRXCKSj+hDqsTbd324D3g+pF+uWKsWLECTzzxBA4fPgwAWL16NdauXYvbb7/dZ98lS5YgJycH999/P1wuF9rb23HHHXfgpptuQlYWOzH+9Kc/xfr16/H++2yg4+HDh/HII4/gnnvuiej4eFkpb83NS5zDJncykJTLImqpjPU8kd0eSpoNmNNQ027F7s4UfOQ+FR/sqYMYwklCjSiKStdj6cJanMHSOJWtviMZeHOnIjmomcZuVUGNSa+Vm9o1RmFUQpvcfdOz1HpmGZMYJ+T30lOjlQc1Oei2K0FN8gCXbgeDN+CraunG9S9+h+Vv7cKrm6W25X7e42jhdLnloYBpglQ2HGxCexRQBzU+n9eEDNYoDoDro18D37L0KC57nlVa+Ju9kzdFHqhYL2TibsdtKF/4TxZIeJOYJbX/F+XPBQ78T5mXBuAczXYlqLF1Ah/9mlVoBRoKGYhyKagpVaUYZknS/v73cPQg661z6ggliEw06vDO7Wfi3dvPRFaCHvjmr7i4+78oFupxsLoV1g0rccnx3wIArKV9T+16oJoa39fme7xnSQ8fFVAgBTV7/wNseIQN0934CPDe7X4bEqpR5j6FH9QkGLSYXpKGZJMOBq0GuamK2pInfV/Vqlq8yY33EtEsnbtSzfqwFj88qOkJGNT471GjZnpxmvy9ViPILSwEQcCrt5yGgw8vwto75+K578/ExrvnI23KQjhELUrFGqApwDDcQx+xgKbxoLLt6KfMwsDnf4Ua1BiTlTTV+gcBt9sjqIlKr5o2HtQUo6Xb7hH/VrR0y9eW7H4o2w5GvwQ11157Le6//34sXboUBQUFePTRR/HBBx9g9OjRqKqqQlFREVatYmMHdDodPvroI+zfvx/FxcWYNGkSpkyZgj/96U/y840ePRoffPABHn74YRQWFmLJkiV44IEH8L3vfS+i4+Mr0Gap+VxGGFG+B4LAJuACrI+Hy4nJbWyu0yc4AwCwXcoVA+wPzechhUqH1Qm703PwF+/MW+lPqZG2Fad7KTU1np2FuerT0BHaFN1gtPlpKQ4A18wqxu+vmIJ7L5wQ/Amk5oZVYjZ67E454h/ozsHB4P+YT3x0CLsq2wAoqph80e6s9ekN0VcaOm1wi4BeKyDBOTBBDQ9OnW7R/2p21k3oSh4FrUv67Cz4DTDxksBPaE4HLvsbcOGTuEj8E/7jOhupgdRRQQAyytj33/6V+dHW/JL9XML+p87R7kRbtxSMb3oG+OZZYPVPgb+eCRz8MCR1AW6X0k+nTDUzKH8qm6ItujHpxMsAgFOK0jweOjY3mfk3dr8JfPQrnLL/CXxpXI5n674H0xePQC+4sFl3KooufbD34wgHD6Wmb833+BBZOagpm8Mm0ANA6VnAmT9nTfl2vc58RkG6ZnvPfeKkBwlqeuSgRgeTXos3fng6XvvhaR49qPgogcrWHqUvl6zUJMn9azLDXJSaJaXEEiCo4eef1GBKjcoTMio70Wd+kVqZzk0x4bFr50DkwXOg4aWf3McWyCPnA+c/JO37qaKImDPC+7ufdRdT9Or2APv/C0EQoturRqUeefc8O9ncjabhpNQAwG233YbDhw+jpqYG3333HebOZRNVi4qKUFVVhauuukret6ioCKtXr0ZNTQ2qqqrw9NNP+/hh5s6diy1btqC6uhpHjhzBj370o4iPjQc1/J8tM8z0kwejz2O3W18EnhiJ5M5jsIk6vNkxBaIoYntFq8fuH+yuDevpuYSXbNLJq6tA6SeXW0S1NFCzhPfw4KmRxgOs0ZhEfio7gdW29z2o8R7+xtFpNbhmdomvsdYbKf1UIeagx+HC/hp28R6T07cTdzThDfjUTd/kTqfGJGVyepTVmtp29vfMTTFBkLt39m/6yazXQq9lJ2W/Iwm0eqwpuQt2UYvPTOcB837Z+5NOvQruWbeiycZOOUEbKPJRFFtfAp4/iwWLGSOBa9+EU2tGntCKlLYDTJnZ/op0TEag6RAzkW7/V+/HU/ENYGtnqlCuV5HA3LvZTfc65KEZpf764dgtzMgPQMwcC6eoQQY60SEm4C77j1G96CVoEsM0Q/ZGapHSeC1K6Se7083S8SkFwE+/Ae7cDdz0IXDBw8C1b7AUxrH1wFdPB3wu3ngv2Uup4akkfwNI5fSTFFxNLkzF7DLPz3VuihFmvRYut6gs4FTpJ34hzQxzUZooKzX+lb1QlJq8VJPcjHF8Xmh/C8MEVqDit1GntZ3N9gKAK14CTpGsDbU7lRRUqCoNJyEDOPNn7PsNjwIuR3S7CrcHDmo8lJrhEtQMZdp7HHC7RSWoiTT9BADjl7K5LuZ0dpIE8Lk4DXU2A2rbrbJSs1hy9H+4u9ZvCsrqcPltJ+4vL8lVGG+lpr7DCodLhF4rII8HEimFbGXvdgIN++R9ubRbFwWlhgeJwU4CQfFKP+2tYe/jlKIwKwT6EbWEeo40r8RjBcpTUFH21dRIfpqCVLM8obu/lRpBEOQUVFuA+U+fOyZiiu1FPG78efBmfCosdqfsgQg66mLR75myM3I+AIGpYBf9CTCnoTmXrXbHdmxiK97OWvZ+LN8HTJMmXu98o/eD2fQMu518uW81YslpEEvnQA8nfqhbg6J0P0HNpmfYa6eVQvjxl/h++mu4wX4PzrE9hc3JF+Diaf4LGPqEIACzb2WDcLkHJkJ4UAOwcw8AVp2ZXqrsNHYhsPB37Hv11HUvelNq/HUu59181cfhjSAIKMtixtLyZimYUQc1ktIe7qKUB1KBZj+1huCpAYBZpSwIm1wYYoA5VuoufHITC2LU1EpVg6nFQGIm69DLVfatL7HbcIMagFUgJmQBLceAna9FWalR0k/NXiN3Kpq7h5enZqjT3uPwNJmFODPEL3oTcMlfgF8eA275FFj8BF5K+SkANtBxn6Q6/OKCsUg0aFHd1oPtUnttTnmTBTMf/hR3r/IdPukvL8mVmoZOm3JCgqLcFKSZ5QFiEASV50N5/nwpqOFKQF+QlRpziO9j01Fg68tspW3tkC/W3CjMU3ST+mO2UYTMH5eNwjQzfrlwHG6ewzrdeqxAZbPwjqi+bp2kpOWnmeRhlkiIsgLgB67k+UtxAuwza4MBHdbQfSx8X4NW45Hf90FnAE65BrhhNXDXfuCnW4AR8wAAlhJWPDCt51v2GQKAad8HkrKBs1ewn6u2+F401DQcAA6vBSCwNIsfGqfdAQC4TrseOZpOzzs7aoCvpfT4+Q8CehNKCgvxhfsUNCEVt5w1ou+VJYE49wHg9s3BW+WHgPr973EEucDx4Kl+T8C0XiBPTTClpkflqQnGiCx2rjveyIMaZZilrNSEuShNlNJP3QF+79YAHkFv7r1wPH65cByuP70stBfOHAVkjmELzGMbPO/jCi8PZABgtNQegp+31QFnqBiTgbmspBvrH8I4Lasi5kpNoLlcIaFKP/EARiddd062WJSgZjh4aoY6bT0OtEiRZYpJF50TkEbL5jmddhsy88sAAO9ur4LTLSI72YhR2Uk4byKrlvhgt2d5+v921cBid+F/u2rkmSMcf9FuWoJelnrVJjofPw1H9tXslDfJSk0U0k+BjMJ+cTmB168GPljGPBNS6qlHlwoLzGjqtONEEzuB8VbbQ4HROcn4+lfn4KcLRiM9kf2eLRbV36qfzMI1UtCZl2rqdSJuNOHGx6MNvl4KURTlv1FYQY0890kXehuBlAI20FHCLaV7x7uPyFUdmHkju00vBTJHs74Zx30HU8pwlWbCUo/nVnM0eTZ2uUfCLNihee9HillWFIF1v2Xeh+LTgImXAlAqoFJMOnzv1AhW1AOMRiPIU5sDGWYBANnj2aRwa3vAgZe9KTXddpfH4gtQAgpzr0FNAKXGmKRSasILahJ6ST+1h6jUFKUn4KcLRvf6O3jA1ZrDXr4afm5Wzxscc4HnPoGmc/fGrJvZbLXuZvzZdh9GC1WwOV3YdLQJk/7fR3jzu4rwn9PtVj4PacWyv4kvRCuaKf00oHT0ONAkVz5F/w0fk8suCOsPsCGPM0rSIAgClk5lzYPW7Kn1mDK84RDbz+kW8dWRJo/n8hfUCIIgVzepfTW8GoobiWX4P4pfpSYaQQ331IRwctn1BpNCAbbabWBO/04zk+u3nWyBKLLj64+/TTTIUK1A5VRi3hQAAktJdATwTdXsBF5azAZEhggv5y5IUSs1gxvUNHTaZE9El80Z8mpPDmr6UNWWlF2KPe4yaFjXdJaiylQ1UBslzS7ifXO8aa9mDf4AYM6ygK9T1WrFrx23wiqY2Kr6w18wT9p7twO732I7LfydnHq7+JQCzBmdiYcvnex/OO4QxCz7K4IENTqj4her2+t3F97awNtTk2zUyd4sb7NwqEpNWSYLangQ7ddTE+Z5ovf0E0+n90HBD8TYRez2yCfMrM7hi6GC6cq2ollsgCUnkvQTwLIJN6wG8qYgXWzHm4ZHoG06hI/31cHqcOO7Ey29P4c3XfWAy87M5MkFciXa9BKmItd2WFEtXY8oqBkA2nsc8j9ZxJVPQeBTXJ3SyX6G9IeeNzYLKSYd6jts+PoYC16au2zYKVXTAMq0a06gvGSJVNZd0exHqcnw6tzLlZqG/YCT/d55KWyfqCg1oXpqnDbg89+z7wUtYGkEPmM5e0sCa9pYLv0+Q0ml8Yaf7Ng0X2m1Z0xSpHqeA1fTdAT49+VsyvJXfwz5tXh6sCjJzQarAv1uFAaUqcdHG32DGjkVINEVolrD36vkYH6aXkhL0GODW3Xin3mT5w6jpaDm6AbfdInbBXz9R8DtYBU+fNq6Hypbu7FPLMNbZQ8yT8/2fwHPzGDVQIIWWLzS4/GZSUa8duvpuKQ/vDT9hDlU0ygvJa/f4/duHtR4KzWCIMj/K95BDffUBO3ZApVS08SNwqqS7gg9kYlS8OUv7WZ1uOTtqZF6BINRcjpgTGWqa/U26UU7lE7R+arPtkarVNcCkQc1ADMN3/A+jutGIUvowJgtD+CANF3cFWabEQBK6imlENDqZJFgXF4yEgxaiKISHFJQMwC0dzvkf4j+CWo8J7/OKGVBjVGnlU96b37HTFafH26EKConmM8ONXioOLxHiXdekvtq1L1qAqaf0kpZxO+ysyooKEpNs8UefKXWCy63KBuFez0JbH+FlScm5QGLpeCm5TgAwJpU5LHr5IKhG9SY9Fp5hdmqPlnPuZPdfvuCp6ejvRp49TIlfVS3R1lx9gJX0oqMUvCpNQKG/usmzOFKzbEGi8fnEVCtmiW8U6aBUJSayJUMk16Lz4RTAQDOxDzWZExN2VmA1sAqM/iF4vDHwCuXAI+XAN/9jW07a1nQ16mS/q+6S88HFj0u/QLVrCDg+neB0yKvvhwqmLzLugORJwU1AZSaTrn5nu//f6BeNerme8HgQU11Ww87T6lLursia8nBX9Ni8w3GeSpdqxH6p6WEVg+MkSpmD37Ibuu8TMJqxqjGrvQlqAGAhAz8Jet+9lTNO3GilnXGjshXw8vMpeaS/G+RlWRUGr9C6tDQD9fYYMRnUNPjkP8I4UqXoVCamSjLrnqtIHfWBYDvnco+BJ/sr0Nzl01WZm44sxRJRhbxqnvZBFJqiv2mn7o97pMRBB9fTVqCXjYLNnRE3oCv0+qQF8RBjcL2btapFADm3c18EKocsT3Z8x825IqCQYKvQD1MkOMvArInsCq4b6WLZ2c9C2jaK5nfIzGHeT74Ki0Idqdbzkvn6FTdhPs6syUESjMToNMI6HG4ZF8PR/Y3SPjtPOwHHvwErXwKgVrzOFxn/zWOXfgmu0ioMSSy1TAAHF0PVH7HyryPf8ZW+YZk4LSfKK0YAuChep52G3Deg6zS8Ycbpaqs2Mfs3YAvELJSEzz95C/tlhHALBxq+ikj0YBkKbg42dztkX7igVK4aWqu1Nh4ObsKpehB3/fZSIGYcDG73fkaU6+5n0ZtEuaMWciqlwpmRGUx024sRJ2YDo3oxEgba47r7otSk8quZ02qRrbqoCYjwTDgXeHjNqhRetREP4rUazXyCmNiQarcGwAAJhWkYmpRKhwuEW9trcQXh1m0fMHEPMwdw7qxqlNQTQHMVt4N+KwOF+ql4KTEO6gBfHw1giBEpQKKS4yJBm1ww/Vnj7E8bFoJMOMH7GKk6m/iTPEMaqYM4fQTEOBkrdGwgA1gDeHq9gIvLWT9U5ILgOv/qzR6q/i219eo77BCFAGDToM0cWAa73HUn2FvX413+qkz1PSTNJKjL54agAXkm9yTUa8v8r8D99XsexdYdSOrNhm3BPjxV8CvTgKLH+81MOQLBLmc+6xlwPdeAzJG9OnYhxIm71EJgciT+vg0H/OrMHbJHcB9g5pAZd3efWoCIQgCRmapfDXS6zt1CfK5J1wlQB1IeQd0gXpuRZXxS9j5wNLI/HWyn2aa776JmcDPtgI3fhiVlzYZdNjqZh6pmZpDACJUalSVT6IoyqbtrCSjR1+ngU49AXEa1LT1OCIuBwwV7quZ4TUjBAC+N5tdwP+y4Sg6rE5kJBowrTgNC6T+Jxu5cdjlltNkvp4aJagRRVGWyxMNWv/eFj8zoKLRqyYkk/AXTwKb/sy+P/f/sZJdADjle0zZ0CfClaN0S81ONiKnt4Z9g4zSLdVLpZh0GSvb7GkF/nY20HqCKVI3fcgCumJJRagM3PeDw1NP+akmCLxNej/PfVITyCx8oon9zOOCkNNPslLTN1lf7qETSCHivprKb1nKKHMMcPkL7OKs6b1SRb1AKE4PMFl+GBCSURhgPVMScwCIrBzei2BKTaCybl791JunBoDcq4YFNeyz1+5izysI4Rt6jTqNPFu32ysF1d6fJmGOVq+MMPjmryqlZrr//c3pgMHPQjUCjDoNtklBzSwNU2pcvViq/KJKP1nsLtmXleml1FBQM0B09Dj6PiKhF3589igsmpSHW87yXdldPK0ACQatvFo5e2w2tBoB88dlAwB2V7WjodOKFgubp6ERfBtMFaaxk63F7kKLxY6t5czBXpaV6F825SXHdXvl8tRodBWWu28mBljZbHoG2PAw+/78h4ApVyr3afXALZ8AP98BXYoymHSoqzQAkCEFjq3eLeA1WqUvhNvJGqXd/LEyqLFEmnpduYWVRQaBK2j56nLuATAJc2Rfjcos7HS55ZTnaMlM3BFq+ikK1U+AsopuD9AYUJ7JBrCOuFe/ElZ7+RqpK3eCQTvgfoCBxGf+UzBkX42nWdjlFgMahQElOGj2qX7iRuHeg0zFLKwoNa0O9hnISDAoPblCRBAEOZjyHgPSGmDkS9SZeSPzx9XuVDoJ+1NqooxRr8VW9zh2CJrDEOCOMP2karwnZRMSDFokGHQoyVTSZAPdowaI06BGnX7qr7LhyYWpeP76mX67kSYZdbhIKu8GICs0OckmTJW66H52qFEeZJmZZPT5xzXptXLX4MrWHrwh9Rq46JQC+CVjJHPdu2zA/5YBX/8JswS26upLBVRbT5DGe9tfYfNMADYfiBtpPX6RFCA518MwOHkINd0LBD/ptfi7sE65Cph0OZuHdOOHQEq+cl/uFECfyHw3jb6rXjWKUjNw3YTV+FNqatpY12qjToOxeSxQCDn9FCVPDb9QBpzULQjA1GtYldJFfwZyJ4b1/HJrhPSE/vNVDAHMvQx29CCAr0atwKQ5m4EXzgb+p/yfcyXcO/gPNf0EKEFNRWMbq1wD0GRnzxtp0CmXdXv1qhmQ9BPABrdOUUYFMZOwn2GwUcao0+CAWIJu0YhUoRujhZrw00+i6JF+avLKepSSUjPwdNmcsrw8WCsxbhjWagScPSZb3r5gHAtwnv/smHwxCRTtcplv7d5a7Kpqh14r4MqZAXwGgqCoBDv/DXz6AK7dfzuKhfq+eWosARrvHV3HgieADVbj3V4DoF6xDeVybo7sqfEzrA9aHXDVy0wh8E4XaXVA0Uz2fZDW8wBQ2+ZHqRnA9JNc1q0Kao5LqaeyzEQ5DRR69RP31PQx/ZTQS/oJYKrgiuPA1KsC7xMAedL9ME49AYCZN98LSamRfDVeFVDy4tAM6N65iSkP2/7FypQRTKkJzSgMKEFNQ3OzvK3Rxh4XqX0gMUBA1yY33uvnoAbwrKDzZxLuB0x6LZzQYaeb9XaapTkUtlJTW1sFOKVrRmqRUnQjZRMK0sxyeo+CmgGEV2z0l6emN6aXpOOxy6fgmWune5RCX39GKQpSTTjeZMF977ETSKAPRpHUj+afX5cDABZOyguuPF3yLLDkKeCs5UBqCTRw43TNgT4qNX6Cmro9wNs/YFU+U7/HWrv3godSEwNBTbAJxL0i+2qCm4Vr5BEJ5gFtvMcZlZ0EQWCSPD9x8XLuEVmJchqJByu9ES2lhquC/trvywgCYE6L6PkDVhEOM0L21AAqpWafR9qUexN/o3tN5RMTgZrtAFSemgBKTYI+dE+NtUuqCtWZ0NTNHh9p9ap5sNNPAAtkpMnzKJzZ/68HZTzGFpGloGZpDoWt1Hy3kxWbNCAdbo3Bo/IJYIUNBZI9goKaASDR6Lky6FdDWC9ce2oJLpyS77EtK8mIF2+cjSSjTs5VB/pgcKXG5mQnmet6a8+elMMG4p33WzbID8BM4XAfPTVeLcW7W4DXrmaGvrK5wMXPhFSCnJVkxOkjM3D+xFy5Kmsok+GvpDtUuGLWi1LDg82CAR6RwDEbtLJ3i6s1clCTnSgbfjvDNQpHzVMT2uuGCzfdD3elxhRO+ilrDOv/Y++UR5sALKi/XPMFLnNI1TkZUnfnSjZZ2l/w73aLsjoUSvopxaRHZqIBCYLUeqIPE7o5XKnp9ko/+ZzP+ptL/8qmwp/6wwF5Oe6jks3CwuGwg5raStZbrMadgeNNFo8eNZzzJuQi0aDF9OL+n1PnTdwFNeoTalqCfsBr6ENhQn4Knv3+DNlHE0h9UbvMyzITcMaoMC54Ui+PWZrDaOyywRGRBV7xNfBUBLa8CHTWsJPbNa8qlU69oNEIePNHZ+DvN8yKCR8DN0a3RnJhLZoNQGAXh+ptwKf/D3jtKlYJ0dUo71brMfdJUmoG0CgMqHw1jV5BTVYikk2RpZ9S+1j9lB5K+qkPVMnpp/hQakJKP2n1bA4U4DG01d50HI/qpQ7aZ/9KuThXsaBGXf3EmzhaVROiQ0k/Aezzlghp8aXuJhzmhG6OWQ5qvNNPIXZHjxYZI4Bz7w/LyN4XuFKzwz0Gbggo1TQgxdncy6MURFFEYwMbA9MiJmN7Ravfzs6/vXgSdjxwAUoyB/5/aOhd0fsZtfQ9lCsbzh6bjSevmorxecm4cEqe333U8vi1p5aEFwwUsa6sozU1SBU7ZVMyTm5mHXBDpFW9snHagS3/kH6Be1gp4jAlqKemN0ypQK5Uwv73c1jr/iOfAB/9CnhqHPD2DbB2tcmyboGHUXiAgxovXw0PakZmJcremFDST263KCs6fVVqUs3cKBzBex8Cygy14a3UhBXUAKxbMwCsfwiwdQKiiBm7H4RZsONY4nT2Py+dV1C1BRBFOY3jFpXgVx1I9NZRWH7prEQkCDyoUXUTjtA+kCDPfwpkFB6614a+wJWaTiSg2sAqc8fY9oX8+KrWHmitrL1EK5Kwo6JV7qXmHWD226T6Xoi/oEZlUuyPxnvR5LLpRfho2TxMLUrze//IrEToNAIMOk1gg3AgEjNZ/w4AMzRHUNfeA5z4Anh5EfBu6FJou9pTs3810FXHymknXRbe8cQY6vST9xiBkOAXCICl6c65HyicxXxI+1fD+fZN0MKFZJOOvbeDYBQGPCugrA4XqiXzcllWorxA6LT1rphY7E7wt6mvnprsZPbe17VblYGiUcJic8qpkmHvqTGE4akBWNCSWsx6L639FbD7LZS2fQubqMeGMb9mzSfzprBS5Z4WoOU4DDqNPOiSr+h5usuk10ATYjn26JwkD6VGNihHeA7nKvfH++o9tvfaoiLGMaoCjdZMNqtunGN/yI/fdrIV6YLUK0hMwraTqqBmkPyp3sRfUGNWBzVDcwp0qGQmGfHSjbPx+q2nRWaYK2bejpmaw6hrtwFbX2bba3f7DgMMgMfK5tu/so2zbgk57RSr+FuBhsX8XwELHwNu+wK48QPWifiH64GbPgJ0ZiRVbMD9ulcxMjsJgqMHcEon9AH01ABKULPtZCvOefIziCLrHJuZaAhLqeHDLA1ajceJNRKKMxKg1Qiw2JUmedGC+2lSzfo+K0pDnZA7CnPMacBlL7ABnzv/DXywHADwJ+fl0GSxBRJ0BqXfSuV3ABQ1hauaskk4hMZ7nNNHZiJBCmpEQ5Iq5RHZOfyGM8qg0wj46mgTtlcw5WFXZZvS7HQQ+qsMBPxvrtMI8vl/fBhBzdaTLUgDC2paxSQcaeiSh40Olfcs/oIa1YkqUulyKDFvbDZmlUW4epcMq7M0h9HSVAcc/IBtt3eyjrghwFc2uR27mT9EawBm3dTLo2If9Qo0ogooczpwxu2+pZylZ7DutwBu1H2CZe5XWLt/gL23hiQMJGNykqHVCOi2u1DTboVeK+CGM0ohCIJS/RRCUCc33jPr+uyZMuq08kr7uJ8p4uHSbXfit+/vw3OfHcU3x5kiNtxNwkAE6SeAjfk46y72vaMbFfoR+JtriafqXTSb3VaxoMa7rJubc0NNPQGsIWemgX2G2l0GWR2I1EJQnJGAy6az4cLPrD8Ch8uNX73LGgteMq2gX2YCDgX4KIsxucnozmN/p1HOY2w2XwhsLVeUGiRkQBSVjvRD5T3rhzGkQ5tUVa40Uuly2CBF6qcIx9Bx/F02xZvTWt5rqsPpcsuN17L3SSrPlKtYlVUckJaoR6fNGVkFVDAmXoL/5dyGixpewILmN4HVb7Lt5oEZZqkmNUGPp6+ZhqP1nTh1RCZmlKbJK2w5/WR1QhTFoMFKtLoJc0ZlJ+JEkwXHGrtw5ujgTcvaexzosbvksSDevPldJf65qdxjm8+k+2GIEtSEWSQw/1fAyU1AzXY8abwDzk6dXOUEACg+FdgMX7OwV/opVJMwwPp5TcjQAC1ATbdWPu9k9WFhevuC0XhnexU2HmrEPf/ZjQO1HUhL0OP+peE1a4wlTh+ZiZ+fMxpzx2bDZnehTkxHntDKSvDVKXE/dFgdOFTfiXR9JwAgMzsP6FTuHyrpp7gLatSemqFsFB4QMsfAqk+FydGOOTUved7XdhIonBH04dxPk4xuGA7/j2087cf9caRDkowEAypbenznP0WBf7gvxrcOO35WchK5aGHD72bcEPXXCYWLA3Sp5gGKyy2i2+6Spx/7g6efkvvop+GMzE4CDjTgmNdwTU6P3YX3dlZj7d46bDraBKdbxGkjMnDDGWW4YFKuR9XjF0dYxVlRuhk1bT1wi8CUoqHfK6mvcE+NLRylBmCVUDesBhwWfPOHbQBs/pWa+n2ArUsZaumTfgo9qAGAMeksqDnSxlLjOo3QpyB5RFYiLjqlAKt31uDdHaw44tcXTui3LvNDAYNOg7suYD1qNh1rwlb3WCzVfsvaS/CgpvkYq3CbfIXHImpHRRtEEcjWdgMiUJBfCLDqbmgimMHVX8RfUKM6qQ4VuWzQ0GjQkTkdprrPkODuYumNEfNYN+DWk70+nJczn2Kqg+B2Asn5QP7U/j7qIUN6XyqggiCKIo43WrDLdT5uuHQecnMHptwzXEx6DXQaAU63iA6rI3hQIys10TnljMpmDdmOBUg/PbB6L1Ztq5J/FgTg2xMt+PZEC04pSsWqH58Jg04Dm9Mlp5z+8YNZyEoy4lBdJ2aWDt/KPU5Ys5+80RkgavWySumxQEwpAFKKgI4qoGYHMqX2/zxNG06PGjWlySyYqe1hj0tPNIRsNA7EHQtG4/1dNRBF4PSRGbgq3IKLGEYrCNjGgxreCFQUgTeuBZoOMQ/V6PPk/bdJ8wUztRbACYwoLgbQBoD9/cOdwdVfxLWnZqhXPw0EzsLZyg/jlyiDL/lsjyC0S3OfJhrYVHFkjo7y0Q1t+tSALwiNnTZ02pzQCEDpIPR5CBVBEDxSUMGIVjdhDh/hcDyAUrP1JPOE3TSnDBt+cTY2/eoc/Oyc0Ug0aLGrqh2fH2bqzNbyVlgdbmQnGzEuNxlZSUbMGZ0lX/CHM+ZwjcJedFidcLhYoOGjehcrvhqelmjxST+FF+CmaSWlR2RpxGicv8fkJuOmM0egJCMBj10+NSZ6ZEULrUbAFmm4JSq/ZZ2iq7aygAZg7T1UsP8pEUkuNgKjrKQIJmnUxlAquom/oIaUGg+MI+fI3zumfh9IL2U/tIWg1EhplzHaOraBV0DECXK31CgHNTylUpyRAKNuaF9clQqo4Ck4Ze5TdIOa6rYen66wFpsT5c3sPfzpgtEYmZ2E/FQzfnHBOFx3Guu6/Y6k4vDU09wxWXF1QQNUAy0jUWqgKJQJBq1vEMj71RzdgGwpqGnstAHdLcip+BBauMJWagQH+5tawM7b0fJwPHDRRHyxYoE8Yype0GgEHBBL0Q0TYG0HGg8Cu15XdqjeJn/rdLmxs7INCbBBK7L/dX1iptxuZKj4aYB4DGrIU+NBxtjTcQyF2OEejUMJM4E0KagJIf3EO7qWoYZtyIyvoKZPDfiCwFMq/MI9lAm1q7Ci1EQn/ZSeaJC7vvKGgJyDdZ0s959s9PFHXCGlF9YfrEdbtx1fHm4CwJpdxhthzX7yA/fI+D2PjlsEaPTAya8wvvUzAEBbRyfwysWYv+ceXKvdgIRw1TA7+zt3gys1tCjtC1pBgAta7NdI5+0TXwB731F2qNkuz/n6+lgzuu0ulJql1hJaA2BIlNO0gzHjKRBxF9Twdv6CMICtsIcwgt6MB4pewuX232JvbZdKqanwGFznD97RtcgleReyxvbnoQ45uDEu2kZhnlIZGQMrR2X+U/D0U3uUq58AJejzNgsfqGXy+MT8FJ/HjM9LwcT8FDhcIl7+uhz7pX3n9FJBNRzhQY3DJUY0JqVFHlXgJ6jJGAnMZaXfo7c9hGR044bOv7FhtwAWaraEbRSGnQX7Fp5+GkLqQCzCPTC7IKWgvvoDU2ySCwCdpN60MCfwG98yO8Kl46RWB1Il5v+dXoolU/PxgzPLBvrwAxJ3QU1xegLOHZ+D/zutFLohOPdpMJhckAoRGuytaQdSCllzLZcN6KoP+ri2bgc0cCPLISk1WXHmqZHnP/WTUpMz9JUaZVK3b2BX32HF3up2j/uj5akBVEFNg6dZmAc1E/wENYCi1jz32VEAwKSClGFd8RIIo145/0Wi1rRYeukVc9ZdQMYo6Lsb8KrhMVwtfizfdZrmAFK1YTZOlJSa1NQ0AOSJ7Cs8qNkBaaYXP9+fco3SP6t6Gxo6rFh3gN134WipLYLU7qMwzYxnr5uBGSVDx1gfd1d1jUbAizfOxsOXTh7sQxkyTC5k5at7qjtYuWaKVAHQi6+mtduOQqERetHOWqOnFvf3oQ4puFLTX+mnWFBqeDOvDj9KzY0vb8HSZ77C54cbVRO6o1dwOSqHvT/HvdJP++Wgxn/V2MWnFECrEWST69wx8Zd6Ajxb5lvD7VUDJf2UHii40JuApU8DAKZpjgEAumb/DE2GIhgEF8Zatvl/XCCkoObMCaXQawWcFad/t2jBg5rdGANA5Sc75TqgcCb7vnobVm2rgtMtYmZpOopNrOP2QA/WDYe4C2oIX3hQc6C2g8nQ6hRUENp6HBglsImtyBwFaIa2qTXaZPSDUVg9XymWlZrKlm5ZMbnvvT3ywNRoKjUjs3yVGrdbxKE61hFsUoF/pSY72Yj5Kg/NvLHxl3oCWPWaTrqwOXtJNfujNVj6iTPybOCUawEAW9xjUT51OfYlsqafY9q+Cu8FpfTTohmjcOjhxZhWnBb2MRMKGskY3y4mKAN2C2cC2WPloEas2oo3t7DrwPdmFwPdfLDu0FFmvKGghkBpRgKSjTrYnW42jTlEs3Bbtx0j5aAmvlJPgDL/qb3HAVckQy39cKLJAlFkikYsyOs8SPFWajYda5K/r2zpkX1CUfXUSEHf8aYueajoyZZudNtdMOo0KMsMrHRdPoOpkWa9Ni560gRCp5WCGlf4n1/FKNxL6m7p03g6cRlusd+NRosL242sMqqk5etefXse2FhQIxiS+tyfhlCUGrdbBCZcxDby5qk8qKnbg/qWDiSbdFg6tUAZn0NKDTGU0WgETJRWtXur21VKTXnQx7V1OzBK4H6a+Kp8AqTJ5GD9qtp7KWkOFX7xH5WTFBMlxkr6yfP3//ooa2jnvZpOjaJSU5xuhl4rwOpwo1aaP8PVoXF5yUE9cwsn5eK2eSPx2OVThnzZfH+i17D3KOpGYY8XMWN31hJ0IAmNnTbs1k1Cl2iC2dYE1O0K/QWl9BMMQz8tGwtopfOLSxSBuXcDd+4Cpl7N7kwvA8wZ0LjtGC9U4LLphawEX1ZqKKghhjg8BbW3uj0MpcahUmriL6jRazWyRyTYUMttJ1tQ124N6TkVP83QTz0B/tNPoijKSs29i8fj/Im5yv5RKukGAJ1WUWN4Cmp/jeSnyfOfelI/9t4LJ+BSaahhvCIrNREojS3BSrq94CW/jV02dDi0+Mo9hd1x+OMgj1LhcrDiBYCCmighxbNMZdbqWCDDEQRZrTlFcwxX8k7LPVJQYx666iYFNQQANgUXAPbWdABprEFZMKOww+VGQ6cVIzXxq9QAql41AXw1Rxu6cMVfN+PH/w7NFHlcrnyKjRO3v47Ch+o70dRlh1mvxfSSdDx48SSkmHTISjJGfT7MSK9xCXI5dwA/DeEJV7MiUWqau3oxCquQg5pOG7rtLmxwT2N3hBrU2FVm8AGeVD9ckdNPov+A1lXAZv9N0xxDSYbU2TwG0k9xN/uJ8M/kQnYR2F/TAVfqFGgBoL0acDlZFO/FyeZuGF0W5Orb2IY49NQA7IRe3twtn+C9OVzfKd/2NskaUHquxI5S45t++uoIU2lOHZEBg06DgjQz1t11NiDAY5BkNGBl3fU+QU2gcm7CE70mck8ND+RD8X5lJylBTY/diY2uaYAerMHbvveYp0OjBWydwLGNgKWB+W0EARi7iLWZAFjTN93Q95rFAnL6KYBK1519CpIBnCIcQxKf6xYD6ScKaggAwIisJCQYtOi2u3DcmoQxWiOTezuqmfS763Vg5k1AGivbPtrQhRGCNB4hMZsNP4tDeH+Txi7/PTdqpEqmbrsLnTannK757FAD7ntvL35/xVSPxm+8vT9XIIY6ckfhHkWp2XSM+WnmjM6Ut+WkmPrl9XnwsmprFWaUpKNGSvOND1DOTXjClZpwq5+sDpc8bTsjhCZ4WV5KTSPS0VUwB0k1XwOrfsCa9WWMAk58Dri8Fgg7Xwcue4F9T6mnqKGRlRr4XXB1pE9BMoDRmhrA0QloU1Xpp6Eb1FD6iQDApEjegXVPTaccvKDyO+CVi4EvnwL+tRToZE2YjjV2YSQ3Ccehn4ZTmMY6bPLgxZuaNsVLU6/y1azZU4uq1h6s2VMrb2vvcchpnKJ0c38cbtRROgozpcbhcuPb4zyo6f9S6YWT8nDu+BzYnG7c9TYznRZnmKNaZTWc4Z4aR5hKDa980msFJAeZzs7hSk1Tl00eaNm0+AVmUDWlsc61Rz9lAU3GSKbcTLyUKTM124HyL9gTUeopamhVQYw/saZdk4ZqUVqY1O9ntzGg1FBQQ8goZuEOxSz8/h1MrQGA1nLg35cDPW042tCFkRrpghxnnYTV5KcyBSJwUKNsr+uwqraz7ytauuVtVa3s+8xEQ9gTjAcL7qmxOd2wOlzYVdkGi92FjERDr2bdaGDQafD89TNx6bQCedtAvO5wgVc/hZt+auF+mgRDSFV63FPT0GlDt9S92JSSDZx7P7B8H7D0j8B5DwK3fwP8bDtwzb+Bq/8FTLqMPcHmZ9ktKTVRQ10W7y8F1WVz4ohbMgg3HgDcLjY6ASClhogNeFDz3s5q7O1m38NpZR/g6/8LJOUC9XuB165CQvVXGC9Usn3iWKkpkJSa2jb/1U217T2q79VBDdteqQpqKlvYtlhRaQAgyaADv6Z1Wp346ijz05wxKnPAeonotRr84eppuFGaP3PuhJwBed3hgKzUhJl+au5tRIIXPKjpsjnlC6g8pduYBMy6CThrGZAzAVAHSbNuZrfSDCIKaqKHVqNWanyDmk6rA4dEKahpOAj0tAGQ9hvC1U+xsRwkBoS5Y7KQmWhAs8WO/1UYMVkPuDR6aL/3OlB6BvB/7wL/vBCo+g6P4juAt/eIs0GWanhQUx1Aqan2k34SRVHev7qtBy63CK1GkJWaovSE/jzkqKLRCEgy6tBpdaLD6pDTaWcN8IBIjUbAby+ehF9cMFb2+RC9I3tqwlRqZJNwiEMlk4w6mPQaj3EMIQ20LD4NyJkINEjpDwpqooY6/RRQqZGDmv2Kn8aY4rd4ZKhASg0hk5tiwlf3nIMnrzoFB7MX4lPXDDyTcR8LaAAgbzJw01pYJlyNBjENACBqjcrwsziEe2rqOqw+Jwab04UmlYGYp59aLHbYnOzk7nCJ8vaqVkmpyYgdpQZQetV8tLcOh+u7kGDQ4sIp+YNyLBTQhIdS/RSmUtMVYjdhCUEQZLUGYF6ckCrhBEFRawDy1EQRjertd/lVapw4LKefDqrKuYeuSgNQUEN4YTZoceXMItxy4Vn4oeNurHXM8NwhdxK2THsUp9qexY8S/wThJ5uAlMG5gA0FspON0GkEuNwiGjs9K6Dq2z1/5g34arxSVTwFJQc1MaTUAEpX4b9/yVIEV80simrnYKL/UNJPYXpqQu0mrCJbNQndrA+ji/PUawC9pNCQUhM1PIzCAZSao6LUnNLSCDQdYd8PYZMwQEENEQDZANvum1Y52tAFQIA2f0pcm4QBlpfOlcqVvVNQ3j9zRcZ7e4Uc1PD0U4wpNVIA09bNKqB+IHlbiKGPXk4/hafU8KAmnGaKaqUmLCO8KQWYcqX0wPgcPtofaHsxCndaHeiGCa0GadF6chO7HcImYYCCGiIA+VJapdPqRJfNc1ghb3Q2OgamSA8Egcq6uUmYmynrAwQ1VS3dzGcjKTXFsRbUqFI+C8ZlY2Q2fS5iBV2EzffkYZYhemoA76AmzHlb5z8EzP81cMZPw3scERBBEGRPtr/0U5fUXqIlcRTbcPJrdktKDRGLJBl1cv8J77lFRxsoqFGTn+a/rJv/PF0a6tjUZYfN6ZK36yXpv6KlGx09TnRKwWNhWmyln3hXYQC4+awRg3gkRLjIYxLCrH5qjSj9pDRgNIcb1JjTgPn3KP2ziKjAU1D+/vz8fNSRLKnxrSfYLSk1RKySJ6WgKKgJjlzW7fU+8e62kwpSYNCxf7WGDpsc1PAJ1pWtPaiUUk9ZScbwT/iDDE8/jclJGvCqJ6Jv8MA6YqUmjKAmK1nZN2ylhugXeNuFYEpNT5pXyw5SaohYhQc16l4rzV02tHY7IAh87g4RqKy7Vvq5IM2M3BQmvdd3WOWg5vSRrFtnRUt3zPppAOD8ibnITzXh3gvHh9SIjRg66DSRDbRskkzx6pRSb3gYhWOkueRwR1Fq/BuFAcCeMc7zDlJqiFgl349Sc0RSaYrSzTCFU8EwjCkI0FWYVzkVpJmRn6KoObx3DQ9qGjttOFKvvK+xxpzRWdh877k4Z3zuYB8KESa8+skZRvWT1eGSUxNZSWEENWpPDZ07hgTcLOzfKCx5KbPGAFAtVqikm4hV8lKlC7Gqvb+ceiKVRiZw+okrNSbkSoFPRUu33LtmQn6K7Fv65gSblxRr5dxEbKOMSQhdqeGtCww6jYefqjf6ZBQm+gVNMKOwFLgmJqUAGSqvXAIFNUSM4k+pIT+NLzyoabHY5WF9nVZlOGV+qhl5UvppR0UbANanIz1Bj+IMFsRsLWeNrWJRqSFil0gGWvKJ9NlJxrDSjVke6ScKaoYCXKnxl37i568kow7InqDcEY/ppwMHDmDx4sUoLS1FaWkpHn30UYh+IkGOKIpYuXIlxo0bh8LCQsyfPx/79+/32KetrQ0/+tGPUFJSgsLCQpx77rnYvXt3fxw+IaF4apSghsq5fUkx6dg/PhR1hr9nqWY9Eo06WfXaUcGCl4I0EwRBQLHUPZh3GOZBDkEMBHKfmjCqn7hSkxWGnwYATHqtrOyQUjM00AYzCttY36kkow7IGa/cEW9G4aamJixYsAALFy5EeXk5Nm3ahNdeew0rV64M+JhHHnkEL7/8MjZs2ICqqipccsklOO+889DW1ibvc8UVV6CnpwcHDhxAdXU1LrnkEpx//vlobW2N9q9ASChKDbtQi6KI/TUdAIDxNAlZRhAE+b3igy25aZhvz5Ma9PGqEa7ulHgFMaTUEANJJH1qmlRKTbjwFBQZhYcGGsH/39/hcstzupJNca7UPP/888jOzsayZcsgCAIKCwvx8MMP46mnnoLD4fDZv6enBytXrsTDDz+MwsJCCIKA5cuXIysrC6+88goAoL6+Hu3t7XjhhReQmMjaZP/85z+H0+nEl19+Ge1fgZDg5tbWbgesDhcaO21ottihEYBxecmDfHRDiwKvBnw8uOGN+bjqxeHbvZUZvp0gBgK5T0046acIKp84/DGk1AwNeFDrPaWbl3MDXKmRghqNHjAO7XN/1IOaDRs24KKLLvLYtnTpUjQ2NmL79u0++2/ZsgVdXV1YsmSJz2PWrl0LAMjNzcXWrVuRkKBcACoqKtDe3o6UFP+Kgc1mQ0dHh8cXER4pZp08o6Wu3Yr9tew9HJmdRJVPXniXdfMyeN6YzzuoKfAT1GQnG+l9JQYUuU9NBOmnSIKaEVlsUcpbHBCDiyZA9RM3CZv1Whb45kwEZvwAmP8rYIi3bYh6UFNTU4OCggKPbUajEZmZmaiurva7f0ZGBkwmr5VsYaHf/QHg2LFjWLx4MebNm4ezzz7b7z6PPfYYUlNT5a/iYupEGS4eaRVVUDMxn1JP3niXdSvpJxa85CQbPc4F/tJPlHoiBhrepyackm4l/RR64z3O3ReMwzPXTh+0Ke6EJ9oASo1sEubVbRoNcPGfgXl3D+jxRUJYQU1jYyOKiooCfq1cuRJ6vR4aje/TBnLJh7v/22+/jRkzZuCMM87Ahx9+GHC/e++9F+3t7fJXZWVlGL8pwZG7Cnf0yH6aiQUU1HjjXdbtnX7SazUe1R8FkoKjTjcVUzk3McDIfWoiKOmORKnJTDLiolMKYNSRIjkU4M33vP/8XKnhLSdiibCOODs7G1VVVUH32bBhA2pqajy22Ww2tLa2orCw0Gf/oqIiNDU1wW63w2BQIv/a2lqf/X/729/ihRdewBtvvIELL7ww6HEYjUYYjSRx9hV1BdQBSamZQEqND96eGl4Fla9KO+WlmOQLAg9mTHot8lJMqOuwklJDDDiRjEmQS7ojCGqIoUWg9FOnVap8CqMP0VAh6umnRYsWYc2aNR7bNm7ciIyMDMyYMcNn/xkzZiA7OxsfffSRx/Z169Zh8eLF8s/PPPMM3nrrLWzbtq3XgIaIHvyifLzRguNNFgCUfvJHocpTs7e6XVZsClRKTG6KKsBRBTs8BUWN94iBRh6TEGL6SRRFNHWyCr5wugkTQxN5TIK3UZgrNRTUADfccANqa2vx3HPPAWCKy4oVK7B8+XLo9Xqf/fV6PZYtW4b77rsPdXV1AIBnn30W5eXluP766wEwU/BvfvMbvP/++z5+HaJ/4f1VvjjcCFFkqzNaofmSm8reE5vTjaXPfAW7041Eg9YjeOEBYnay0UN+v33BKCyZmo/Fk/MG9qCJuEcfZvrJYnehx8EaTFJQE/sEVmpUjfdijKgfcXp6OtatW4c77rgDjzzyCAwGA2699VasWLFC3ueqq64CAKxatQoAcM8998DlcuH000+H3W7HuHHjsG7dOqSns3bMGzduhNVqxYIFC3xe7+qrr8Yf/vCHaP8ahES+pC40dCqt/QlfjDotSjISUNHSDb1WwIVT8vHDuSPl5maAos4UeJVtzx+Xg/njcgb0eAkCCL+km6dPEw1aJMbgBY/whJ+evJvvcaUmyegrRAx1+uVTOWnSJGzcuDHg/TyY4Wg0Gtx333247777/O7/gx/8AD/4wQ+ieoxEaHiXIlPqKTB/uW469lS344KJeX7VLG6wnlJI7yExNJCb74VY0s0rn8LtJkwMTQJN6eaemlhMP8XeERMDSr53UEOVTwGZWpSGqUVpAe+fPzYbHy+bh7Is8s4QQwN5TEKYSk0k3YSJoUfAPjXW2PXUxN4REwNKRqIBBp0Gdmk2ESk1kSMIAnViJoYUykDL0JSavpRzE0OPQEbhTlvsempoSjcRFHUDPpNeI3cEJQgi9gm3+V4TlXMPKxSlxnN7l3fzvRiCghqiV/gwxnF5KXIHSoIgYp9wq5/kCd2UfhoWyM33AhqFKaghhiFcqaHUE0EMLyKtfiKlZnggj0kIUNIdi54aCmqIXlk4iVXzXDSV5rUQxHBCH2H1ExmFhwe9DbRMNlFJNzEMWTwlH4tpAB1BDDt0EVY/UUn38EDKPvqkn2K5+R4pNQRBEHGKXP0UglIjiiKautiIBEo/DQ8CpZ+6bNLsJwpqCIIgiFhBrwldqenoccIuGYqzkgy97E3EAho/RmGHyw2rg/2dyVNDEARBxAxKn5reg5rGLjakNcWk85hdRsQu/pQaXs4NkFJDEARBxBBySXcI6acGqnwadvgzCnOTsFmvlT1XsUTsHTFBEAQRFXRhpJ/ITzP8UPrUKNs6Y7jxHkBBDUEQRNwSzpgEpUeNqZc9iVjBb/qJl3PHYOoJoKCGIAgibpEHWoYwJkHpJkwm4eGCP6Mwn9BNSg1BEAQRU+hUngpRDB7Y0Nyn4Qe3zPjz1MRi5RNAQQ1BEETcojaC9lYB1SwFNVmJFNQMF7R+jMKx3HgPoKCGIAgibuHVT0DvFVAt3SwtkZFI6afhgr+gRhlmGXsjEgAKagiCIOIWXv0E9K7UtFiYUpNOQc2wgVc/uf14aij9RBAEQcQUHkpNLxVQrRZSaoYbfvvUxPCEboCCGoIgiLhFEAQ5BRGsAsrmdMlpiYwECmqGC1p/1U828tQQBEEQMQqvgArWq6ZN8tNoNULMruAJX4KNSaCSboIgCCLmkHvVBPHUtFhYN+H0BL2csiBiHyX9pGzrIqWGIAiCiFV0Icx/4kEN+WmGF/6MwhYKagiCIIhYhVdABat+UpQaCmqGE/6Mwnbpc6CPwWGWAAU1BEEQcY08qTtIUNPaTUrNcMSfUdglKXa6GE0zUlBDEAQRx8hDLUNIP1GPmuEFF2PURmEe3OpIqSEIgiBiDb2md6NwK/fUUPppWOEv/cRL+7Wk1BAEQRCxhmwUDlLS3UxG4WGJv/QT/xyoGzPGEhTUEARBxDGyUThI8z3y1AxP/PWp4Z8D9QiNWCI2j5ogCIKICvoQlJoWaUQCeWqGFxpZqVG28VSUjpQagiAIItbghtBgJd3kqRme+FVqXFT9RBAEQcQoOk3w5nuiKKKlm1c/6QfsuIj+x59R2EXpJ4IgCCJW6W1MgsXugt3JAh7y1Awv/BuFKf1EEARBxChyn5oAnhqeejLpNUgwxGbrfMI/fvvUUPM9giAIIlbhaQZngOqnFvLTDFs0XkqN2y2Cfwyo+R5BEAQRc/RW/aT4aSioGW5ovTw16sCW0k8EQRBEzNFb9VNLF/WoGa7I1U8iD2qUwJbSTwRBEETMoe+l+ok33qMJ3cMPOf0kKTTqwJaqnwiCIIiYQzEK9+KpIaVm2KH0qWE/q0u7SakhCIIgYg5dLyXdNCJh+MKVGq7ScV+VRlB62MQaFNQQBEHEMb2ln7hSQ0bh4YdsFJbiWWeMN94DKKghCIKIa3o1ClNJ97DFu09NrDfeAyioIQiCiGt0vZV0W2hEwnBFKykySkl3bDfeAyioIQiCiGv0vTTfa+1mE7rJUzP84GMSlJJurtTEbmgQu0dOEARB9JlgYxJcbhFtZBQetnDrjFLSTUoNQRAEEcMEG2jZ0eOQ2+ZTn5rhh/dAS2VCNwU1BEEQRAzCL2AOP9VPzZKfJtmkk4MfYvig9KnxbL5H6SeCIAgiJgnWp4Z61AxvNBpSagiCIIhhhDzQ0o9SI1c+UeppWCIbhaU/Pa+Ao5JuLw4cOIDFixejtLQUpaWlePTRRyGK/p31ACCKIlauXIlx48ahsLAQ8+fPx/79+wPu/8c//hGCIOCzzz7rh6MnCIKIH3ijNX99alqloCaTlJphSaAp3dR8T0VTUxMWLFiAhQsXory8HJs2bcJrr72GlStXBnzMI488gpdffhkbNmxAVVUVLrnkEpx33nloa2vz2XfPnj3485//jKysrGgfOkEQRNzhr0/N/poOPPnxITz32TEA1E14uKLxMgrLfWpIqVF4/vnnkZ2djWXLlkEQBBQWFuLhhx/GU089BYfD4bN/T08PVq5ciYcffhiFhYUQBAHLly9HVlYWXnnlFY99rVYrrrvuOjzxxBNITEyM9qETBEHEHUr6iV3YdlS04sI/f4m/bDyKipZu6DQCFozLGcxDJPqJgEZh8tQobNiwARdddJHHtqVLl6KxsRHbt2/32X/Lli3o6urCkiVLfB6zdu1aj22//OUvMWbMGFx55ZW9HofNZkNHR4fHF0EQBOGJkn5iq/TjjRYAQFG6GU9fcwq23ncelkzNH7TjI/oPXuTkaxSm9JNMTU0NCgoKPLYZjUZkZmaiurra7/4ZGRkwmUwe2wsLCz32X7NmDd555x387W9/C+k4HnvsMaSmpspfxcXFEfw2BEEQwxtZqZFW6d12JwBgSmEqLptehDQyCQ9b5PSTd/O9GE4/6cLZubGxEdOnTw94/5133gm9Xg+NnyhPEPy/SaHs39DQgJtvvhkvv/xyyF6ae++9F3fddZf8c0dHBwU2BEEQXshKjXRhs9hdAIAEQ1iXByIG8U4/8eBGG8Ppp7A+tdnZ2aiqqgq6z4YNG1BTU+OxzWazobW1FYWFhT77FxUVoampCXa7HQaDsiKora2V97/ppptwxRVXYPHixSEfq9FohNFoDHl/giCIeMTbKNxtY0pNolE7aMdEDAw+RmFJrYvlRotRP/JFixZhzZo1Hts2btyIjIwMzJgxw2f/GTNmIDs7Gx999JHH9nXr1slBzJo1a/Dcc89BEAT56+TJk1iwYAGKioqi/SsQBEHEDd5jEkipiR8UpYb97BwGSk3Ug5obbrgBtbW1eO655wAwxWXFihVYvnw59Hrf0fV6vR7Lli3Dfffdh7q6OgDAs88+i/Lyclx//fUAWB8b76/S0lJs3LixV+WIIAiCCIz3mATuqUk0kFIz3NFq/Jd06+PFUxMK6enpWLduHe644w488sgjMBgMuPXWW7FixQp5n6uuugoAsGrVKgDAPffcA5fLhdNPPx12ux3jxo3DunXrkJ6eHu3DIwiCIFR4j0mw2CSlxkhKzXDH1ygc+9VP/fKpnTRpEjZu3Bjwfh7McDQaDe677z7cd999Ib9GeXl5pIdHEARBSOi9PTWk1MQN6jST2y3CxZvvUfqJIAiCiEV8qp9IqYkbtKoqY5coqqZ0U1BDEARBxCCk1MQv6iyTyy2qSrpjNzSI3SMnCIIg+oyPp4aqn+IGj/STKMqBbSwbhSmoIQiCiGN8qp+oT03coFGln5xukUq6CYIgiNiG+tTEL95GYR7UUPM9giAIIibRqaZ0i6KoeGpIqRn2eBiF3aIc2FL1E0EQBBGT6FWm0B6HS66AIaVm+KPReFY/OamkmyAIgohl1OW77T0O+fsEqn6KC3SqUQlKSXfshgaxe+QEQRBEn/EX1Bh0mpj2VRCho1GNSuDN98goTBAEQcQk6vRTezcLaqhHTfzAfTVulaeGSroJgiCImESjEcAX5lypIT9N/CAPtfQo6Y7d0CB2j5wgCIKICtxDwYMaqnyKH3hAqzYKk1JDEARBxCx66cpGSk38odX4pp9ieUp37B45QRAEERW4UtNBSk3coVUZhXn6iUq6CYIgiJiFpxtIqYk/+KgEl1uEQ5r9RFO6CYIgiJhF65V+ouqn+EGr6lPjotlPBEEQRKzDPRSyUmMkpSZekJUaUV3SHbuhQeweOUEQBBEVvNNPpNTED54l3dR8jyAIgohxvEu6yVMTP8jpJ1E9pZuCGoIgCCJG0cmeGjahm+Y+xQ9ynxpVSTc13yMIgiBiFr1XSTd5auIHjz41vPkepZ8IgiCIWIWX8Nqlkl7y1MQP/ozCNKWbIAiCiFn0XukG8tTED/5nP5FSQxAEQcQo3s3WqKNw/OBhFHbR7CeCIAgixvFON5BSEz8oHYVBSg1BEAQR+3gbQ0mpiR/8pZ+o+R5BEAQRs/ikn0ipiRu0gm/6iZQagiAIImbxTT+RUhMvcI+4h1JDfWoIgiCIWMU3/URKTbzgL/1EU7oJgiCImEWt1GgEwKijS0O8oBiFlfSTjtJPBEEQRKyiLuFNNOggCLF7USPCgys1TrcbklBDzfcIgiCI2EWn8lAkUOVTXMGNwnanW9lGSg1BEAQRq+i8lBoifuABjE0V1FDzPYIgCCJmUfclIaUmvvAX1JBSQxAEQcQsamModROOLzQa3/QTlXQTBEEQMYvaGEoTuuML7qnhSo1GUAKdWISCGoIgiDhH3acmgXrUxBVaL6VGF8MqDUBBDUEQRNxDSk38opGVGheA2G68B1BQQxAEEfeoq13IUxNf8HiWp59i2SQMUFBDEAQR96iNwjShO77wTj/F8oRugIIagiCIuEedfiKlJr7wTj+RUkMQBEHENJ5jEkipiSd8lBoKagiCIIhYxnNMAik18YTGq6Q7luc+ARTUEARBxD00JiF+8S3pJqWGIAiCiGFoTEL84j0mgUq6CYIgiJjGo/qJlJq4wtcoHNthQWwfPUEQBNFnPJQaMgrHFfxPr5R0k1JDEARBxDAenhoyCscV3rOfqKTbDwcOHMDixYtRWlqK0tJSPProoxBFMeD+oihi5cqVGDduHAoLCzF//nzs37/fZ7+33noLkydPRmFhIUaPHo0nnniiPw6fIAgirlBXP1FJd3zhPaU7lid0A/0Q1DQ1NWHBggVYuHAhysvLsWnTJrz22mtYuXJlwMc88sgjePnll7FhwwZUVVXhkksuwXnnnYe2tjZ5n7fffhvLli3Dv//9b1RXV2Pt2rV48cUXsX379mj/CgRBEHGFx5gEUmriClJqeuH5559HdnY2li1bBkEQUFhYiIcffhhPPfUUHA6Hz/49PT1YuXIlHn74YRQWFkIQBCxfvhxZWVl45ZVXAAButxsrVqzA7373O0ybNg0AMGbMGOzfvx8zZsyI9q9AEAQRV6h7k5j1pNTEE95KDVU/ebFhwwZcdNFFHtuWLl2KxsZGv6rKli1b0NXVhSVLlvg8Zu3atQCA3bt34+TJk7j00ks99tFq6Z+PIAiir/DqJ7NeG/MrdSI85D41Lpr95JeamhoUFBR4bDMajcjMzER1dbXf/TMyMmAymTy2FxYWyvsfPXoU+fn52LVrF8466yyUlpZiwYIF+PzzzwMeh81mQ0dHh8cXQRAE4Ut6ogEAkJ1sHOQjIQYann6Sf47xoDas5GljYyOmT58e8P4777wTer0eGj9GI0Hw/0aFsr/L5UJXVxeefPJJrFq1Cjk5OXj77bdxwQUXYPPmzX5TUI899hgefPDBUH4tgiCIuKYwzYzn/28mCtPMg30oxACj8QpiYr2kO6ygJjs7G1VVVUH32bBhA2pqajy22Ww2tLa2orCw0Gf/oqIiNDU1wW63w2AwyNtra2vl/UtKStDZ2YkXXngB+fn5AIBrr70Wr7zyCt544w2/Qc29996Lu+66S/65o6MDxcXFof+yBEEQccSiyXmDfQjEIOAdw1DzPS8WLVqENWvWeGzbuHEjMjIy/AYfM2bMQHZ2Nj766COP7evWrcPixYsBAFOmTEFaWhrsdrvP441G/3Kp0WhESkqKxxdBEARBEAre6Saa0u3FDTfcgNraWjz33HMAmOKyYsUKLF++HHq93md/vV6PZcuW4b777kNdXR0A4Nlnn0V5eTmuv/56AEBSUhJ+8Ytf4Oabb0ZDQwNEUcR//vMffPHFF7juuuui/SsQBEEQRFzgnX6KK09NKKSnp2PdunW444478Mgjj8BgMODWW2/FihUr5H2uuuoqAMCqVasAAPfccw9cLhdOP/102O12jBs3DuvWrUN6err8mF//+tdwu92YNWsW7HY7ysrKsGbNGkycODHavwJBEARBxAXeU7l1MV79JIjBWv0OIzo6OpCamor29nZKRREEQRAEgFVbK/HL/+yWf77hjFI8dMnkQTwiX8K5fsd2SEYQBEEQRMR4p5tiPf1EQQ1BEARBxCk+RuEYTz/F9tETBEEQBBExmmHWfI+CGoIgCIKIU6ikmyAIgiCIYYGvUhPbYUFsHz1BEARBEBHjrdTQlG6CIAiCIGISb19wrM9+oqCGIAiCIOIUSj8RBEEQBDEs8C3pJqWGIAiCIIgYREsl3QRBEARBDAe8B1rqKf1EEARBEEQsQmMSCIIgCIIYFngbhamkmyAIgiCImMSnTw2lnwiCIAiCiEW8jcKk1BAEQRAEEZN4CzNU0k0QBEEQREziaxSO7bAgto+eIAiCIIiI8U4/0ZRugiAIgiBiEu8+NVTSTRAEQRBETOJrFI7tsCC2j54gCIIgiIjxLekmpYYgCIIgiBjEO/1EJd0EQRAEQcQkPkZhSj8RBEEQBBGL0OwngiAIgiCGBd5BDE3pJgiCIAgiJvFOP2nJU0MQBEEQRCziMyYhxtNPusE+gKGGy+WCw+EY7MMgiJDQ6/XQarWDfRgEQcQow81TQ0GNhCiKqKurQ1tb22AfCkGERVpaGvLy8iAIsX0yIghi4NEMs+Z7FNRI8IAmJycHCQkJdIEghjyiKKK7uxsNDQ0AgPz8/EE+IoIgYg0fo3CMe2ooqAFLOfGAJjMzc7APhyBCxmw2AwAaGhqQk5NDqSiCIMLCxygc4+mn2NaZogT30CQkJAzykRBE+PDPLXnBCIIIF++OwlTSPYyglBMRi9DnliCIvsDVGUHwDXJiDQpqCIIgCCKO4SmoWFdpAApq4orPPvssJlb1giDgs88+G+zDIAiCiAt4LBPrfhqAghqCIAiCiGu4UhPrE7oBCmoIgiAIIq7hPppYn9ANUFATEFEU0W13DviXKIphHWd1dTUuuugiFBUVYfLkybjsssswadIk2Gw2rF69GpMnT0ZBQQHmz5+P3bt3ezz2t7/9La688kr8/ve/x8iRI5GXl4c777wTVqs1pNc+efIkrrjiChQVFaGoqAi33HILOjo68PbbbyM5ORkWi0Xe991330VWVhZsNhucTid+97vfYfTo0SgsLMQZZ5yBzZs3B3yd1atXY9q0aSgoKMDs2bPxxRdfhPUeEQRBEIHhaafhkH6iPjUB6HG4MPGBjwf8dfc/tBAJhtD/LDfccANmzpyJ999/H+Xl5TjzzDOxevVq7N+/H9dccw3ee+89LFq0CCdOnMCll17q8/iPPvoII0aMwKFDh9DS0oJFixbhN7/5DZ566qmgr2uxWDBv3jwsXboUJ06cgMvlwq233orHH38cDz30EP5/e/ceF3O+/wH8NTEzSLVS0cxURlHkltJyXNYqSyiXZZEol3DclnRcfiy55mhZOqfjtptaFudodSzWtVzqKCusu7RUdEGlJGqaxuf3R6fvMabLRNM04/18PHrs9p339/v9vOc903x85vP9fgICAhAVFQVfX18AwL59+zBp0iQIhUIsXLgQR44cwenTpyGVSnH48GH4+Pjgzp07EAqFSuc5evQovL298csvv8DNzQ3Hjh2Dp6cnbt68CWtra7WfJ0IIIZX730Rh3e/U0EiNjktISICvry94PB6kUin69OmDuLg47Nq1C56enhg8eDAAQCqVYs2aNSr7W1paYuPGjeDz+WjVqhWCgoKwc+fOGkeMoqOjUVRUhO+++w58Ph9NmjRBeHg41q9fj8aNG2PmzJmIjIwEABQWFuLYsWOYPn06SktLsW3bNgQHB0MqlQIAhg8fjrt376p0aABgy5Yt8PPzg5ubGwBg6NChGDBgAMLDwz/oeSOEEFKu4usnXV+hG6CRmio15TfCndWDtHLe2nB2dkZkZCSCg4Px8OFDnDt3DjNmzMDp06fRtWtXpVhjY2OV/cVisdIVUe3bt0dRURFyc3Nhbm5e5XnT09Nha2sLgUDAbXv7/6dPn45169YhPT0dsbGxcHV1hYODAzIzM1FcXIwOHTooHe/tfd+WkZGB33//HceOHeO2lZSU0J2fCSGkjujTJd3UqakCj8er1ddA2rJo0SL4+/sjKioKTZo0wcqVK+Hu7o6ffvoJ9+/fV4p98uSJyv55eXlKv6empsLQ0BBmZmbVntfGxgapqamQy+Xg8/nc9rKyMjRu3BgWFhYYPXo09u7diwsXLmDGjBkAAHNzczRt2hTJycno2LEjt9+7x6lgZ2cHb29vrFixouYngxBCSK3p05wa3e+WfeT8/Pywd+9ePHz4EHfu3MGcOXMAAP7+/jh27BiOHDkCALh27RqCgoJU9r99+zY2bdqEN2/eIC8vDytWrMC0adNqvJ/NyJEjYWhoiK+//hoymQwKhQKrVq2Cj48PFzN37lz88MMPuH79Or788ksA5SMys2fPxpIlS/DHH38AAJKSktC+fXtuYca3zZ8/H1u3bsX58+cBlI/SLF26FHv27Kn9k0UIIURFxQCNrq/QDdBIjc5bv349RowYgebNm0MoFMLS0hJz587FhAkTEB0djSVLlmDq1KlwdHTExo0bMXz4cKX9+/Tpg/z8fNjZ2aGkpASjRo1CcHBwjec1NDREXFwcFi5ciLZt28LAwAB9+vRBaGgoF+Pq6gpjY2OMHDlSab7Mhg0bYG5uDg8PDxQXF6N169bYvn07LCwsVM7j7u6OiIgIBAYGIjMzE02aNMGIESO4ThIhhJAPw92nRg9GanisttcQ66jCwkKYmJjgxYsXKnNLSkpKkJqaCqlUiiZNmmiphbVXUFAAFxcXBAcHw8vLC6WlpTh69Ch8fHxQUFAAIyOjavcPCgrCuXPnNHb33pKSElhaWiIxMRH29vYaOQfR3dcvIaRhGLDpHB7mvIKT9SeIntVb281RUd3n97topEaHZWdn48GDBxCJRBAKhRAKhWCMwczMDE2bNv2gY2dkZKBnz56VPubg4IAzZ85Uu39+fj5Wr14NV1dX6tAQQkgDRhOFSYPQoUMH7NixAzNnzkRBQQF4PB66dOmCM2fOoHHjDyutRCJBRkbGe+/v7e2N58+f48CBAx/UDkIIIZqlTxOFqVOj46ZPn47p06e/176VTRyuK8ePH9fYsQkhhNQdA1r7iRBCCCH6oGKERh8mCmukU3P37l14eHjAxsYGNjY2WLduXbV3qGWMISQkBPb29hCLxejfvz/u3LmjFHP16lUMHToUVlZWsLKygoeHB65du6aJ5hNCCCEfDa5ToweXdNd5Brm5ufj8888xaNAgpKWl4eLFi/jpp58QEhJS5T5r167F7t27ERsbi4yMDAwfPhzu7u4oKCgAUL5o42effYZBgwYhPT0daWlpcHd3R//+/ZGZmVnXKRBCCCEfDRqpqcb27dthbm6O+fPng8fjQSwWY82aNdi0aRPkcrlKfHFxMUJCQrBmzRrulv0LFiyAmZkZfvzxRwBAfHw8+Hw+5s2bBwMDAzRq1AgLFy6EgYEBLl26VNcpEEIIIR8N7j41NFKjKjY2Fp6enkrbhg0bhpycHFy9elUl/vLlyygqKsLQoUNV9qmYbNqtWze8fPkSFy5c4B7/z3/+A5lMBmdn57pOgRBCCPloVFzJrQ+rdNf51U9ZWVkQiURK24RCIVq2bFnpV0VZWVkwNTVVuWmYWCzG0aNHAQD29vb4+eef4e/vDycnJzDGkJ+fj3PnzsHGxqbSdshkMshkMu73wsLCD02NEEII0Tv6dEl3rUZqcnJyIJFIqvwJCQkBn8+HQSU38KlqLSF14x89egRjY2O4urrCxcUFaWlpOHXqVJUTkIODg2FiYsL9WFlZ1SZVndamTRtERESoFRsdHY127dpBIpFgwYIF4PF4GrvDcEMSERGBNm3aaP0cEokEBw8e5H6Pi4tD586dIZFIMHr0aI22jxBCgLcv6db9r59qNVJjbm5e4w3ZYmNjkZWVpbRNJpMhPz8fYrFYJV4ikSA3NxelpaUQCATc9uzsbC5+3759CAkJwY0bN2BiYgIAmDZtGrp27QorKyv4+vqqHHfp0qUICAjgfi8sLPyoOjbqKC0thY+PDyIjIzF69GgUFBRgy5Ytau9/+PBhnD9/Hps3b9ZcI/Xcu++nyZMnY+bMmQgMDERBQQFWrFgBe3t7TJgwQUstJIToO5ooXI3Bgwfj119/Vdp29uxZmJqaonv37irx3bt3h7m5OU6cOKG0/cyZM/Dw8ABQPlHYycmJ69AAQIsWLeDi4lLlRGGhUAhjY2OlH6IsOzsbr1+/houLCwDgk08+qdX+165dw/PnzzXQso/XgwcPlOpx4cKFSifYE0JIXWlEN9+r2qRJk5CdnY1//OMfAMo/OBctWoQFCxaAz+erxPP5fMyfPx/Lly/HkydPAABhYWFIS0vDxIkTAQCff/45YmJiuI4PYwy//PILjh8/jkGDBtV1CvjvSYDSV/X/U8v1RTMyMjBixAi0bt0aDg4OCAsLU3o8Li4Of/rTnyASidCpUyccOnQIAJCQkMCt7dSrVy9IJBIoFAqV40dERMDBwQGWlpbo3bs3bt26BQD44YcfsHnzZhw8eBASiQRr164FADx9+hQTJ06EjY0N2rRpg0mTJuHp06fc8YKCgjBu3Djs378fbdu2RUJCQo05RkdHw8nJCWKxGB06dEB4eDgAYMiQIZg8ebJSrJeXF+bMmQMASE9Px5dffsl9PTp16tQq51a9fv0agYGBkEqlsLKywsSJE5GXl1dj2yqeI1tbW7Rq1QoDBw7E3bt3lR6PjIyEo6MjzM3NMWbMGLx+/Zp7rOLrvoyMDEgkEgDAmDFjIJFI4O7ujoSEBAQEBEAikSAmJkat9hBCSG0Y6NFITZ1PFG7RogXOnDmDOXPmYO3atRAIBJg2bRoWLVrExYwZMwYAuLkEixcvhkKhQM+ePVFaWgp7e3ucOXMGLVq04OJlMhlWrlyJqVOnQqFQwNbWFnv27MHw4cPrOoVy8tfAelHNcXXt/7IAgaFaoQqFAqNGjUKnTp3w6NEjMMawbNkyPHr0CED5SMrAgQMRHh4Ob29vJCUlwd3dHVZWVujVqxcSEhIglUqRkJBQ6dyPe/fuYfny5Th58iQcHR0REBCAyZMn4/Lly5g6dSoeP36MtLQ0bv5OcXEx+vTpAzc3N6SkpAAA5s2bhz59+uDGjRvcIpuPHz/GpUuXkJKSgkaNGlWb45EjR+Dt7Y1jx45hwIABSE5OxhdffAFXV1fMnTsXX331Ff7+97/D0NAQz58/x8mTJ3H58mW8evUK/fr1w7Bhw5CamgqFQoFp06Zhw4YNWL9+vcp5fH19kZmZiaSkJBgZGcHf3x9+fn44cuRIte17+fIl/P39kZycDKlUisjISBQXF3OPZ2ZmIj4+HlevXkVxcTG6du2K8PBwruNVoWKtLR6Ph4MHD6J///4AgP79+8PPzw9+fn7VtoMQQt6XPl3SrZG1nxwdHXH27NkqH397YiQAGBgYYPny5Vi+fHmV+/j4+MDHx6fO2qgPrly5gitXriAmJoabjxQSEsJ1MsLCwuDm5gZvb28AgIuLCyZPnoywsDC1JhI7ODggNTWVG2Hz9fXF3/72tyrjDx06hPz8fISGhnLtCQ0NhaWlJaKjo7l23L59GydPnqyxQwMAW7duxbRp0zBgwAAA5VfCpaSkQCAQwNHREZaWlvj5558xadIkHDx4EN27d0eXLl2wd+9eFBUV4bvvvgOfzwefz0d4eLjSvK0KGRkZiIqKwm+//YaWLVsCALZs2QIzMzOkpaVVO9lXIBDAzMwMkZGRCAgIUOl8NGvWDFu3buVWUX97tIsQQhqCijk1dEm3PuM3Kx810cZ51fTw4UOYmZnByMiI28bj8dC8eXMA5R/WiYmJSh/KpaWl6NSpk1rHf/XqFVatWoVTp04hPz8fCoUCZWVlVcanpaWhbdu2Sh0HgUCAtm3bIi0tjdvWuXNnro01SU9Px6hRo5S2VRyfx+Nh1qxZiIiIwKRJk7Bv3z5ucc/09HTY2tqqtKUyFZN1K0YQK5iYmODhw4fVdmqEQiEuXryI1atXw9bWFp6entiyZQs3/6tFixZo1ux/NRUIBCgpKVErd0IIqQ8G3CXdNFKjv3g8tb8G0hZLS0vk5uYiPz+f+6pOJpNxy0vY2dlBJBJxc1BqKzAwEFeuXEFUVBTs7Oxw9+5ddOzYscp4a2trpKamoqysDI0bl7+05HI5UlNTYW1tzcWpM0JTwcbGBsnJyUrb5HI5N3o0efJkrFixAgkJCbh58ybGjh3L7ZeamqoUC0CpbRXs7OwAABcuXFBqp7qkUil2796NwsJCjBs3DkuWLMG2bdtqfRxCCNGGivnBNFGYaFWfPn3QsWNHzJs3DzKZDEVFRfD19eVGU2bNmoWoqChERUWBMYaysjJs3boVwcHBah2/sLAQ1tbWkEqlKCoqwl//+lcA4Ca6NmvWDM+ePQNjDAUFBRg1ahSaN2+OgIAAyOVyyOVyBAQEwMjICCNHjnyvHOfPn49du3bh1KlTAMonIvfu3Zu727SJiQl8fHzg6+sLb29vblRk5MiRMDQ0xNdffw2ZTAaFQoFVq1ZV+hWmmZkZfHx8MGPGDG5S84MHD7g7YVcnOzsbCxYsQE5ODoyNjeHk5IQXL168V66VqXiOX716hdLS0jo7LiGEVDBpylf6ry6jTo0Oa9SoEU6ePIlXr17B2toaTk5OGDx4MPf1UseOHXH8+HGEhoZCJBLB1tYW169f576iqcn69euRl5cHsViMTz/9FGPGjEG7du1w8+ZNAMCoUaO4CbKHDh2CoaEh4uPjkZOTAzs7O9ja2iI3Nxfx8fEwNHy/Ua9hw4bhwIEDWLp0KUQiEfr27Yvx48dzl/sDwPTp05GSkqKUl6GhIeLi4pCbm4u2bduiTZs2uHfvHkJDQys9z44dO9C1a1f06tULVlZWGD16NCZPngxzc/Nq22dqago+n49u3bpBIpEgLi6u0onI78vf3x8hISFwdXXF7du36+y4hBBSYdbndlg93BEjnVTvJadreKyqW/LqmcLCQpiYmODFixcq96wpKSlBamoqpFKpynINpOHbv38/QkND1bo8XB/R65cQos+q+/x+F82pIVq1efPmKu9I/M0332DGjBlV7vvmzRukpKRgzZo1+OabbzTVRO7+Me8yNDRUme9DCCFEe6hTQ7QqICBAaTmL2sjOzoazszOmTZuGcePG1XHL/qempUEIIYQ0DNSpITpLLBajqKhI280ghBDSQNBEYUIIIYToBerUEEIIIUQvUKfmLW/evNF2EwipNXrdEkJIOZpTg/Jb1xsYGCArKwvm5uYQCATg8XT/zopEvzHGUFpaipycHBgYGFS5DAQhhHwsqFOD8gU1pVIpsrOzkZWlhfWeCPkAzZo1g7W1NQz0YN0WQgj5ENSp+S+BQABra2uUlZVBoVBouzmEqKVRo0Zo3LgxjSwSQgioU6OEx+OBz+crLYBICCGEEN1A49WEEEII0QvUqSGEEEKIXqBODSGEEEL0wkczp6ZiMfLCwkItt4QQQggh6qr43K74HK/OR9OpycvLAwBYWVlpuSWEEEIIqa2XL1/CxMSk2piPplNjamoKAHj06FGNTwoA9OjRA5cvX1b7+LWJ11RsYWEhrKys8PjxYxgbG2utHZo6dm3z01Q7ahtbm3hN1rC28ZTj+8fSe7F+2lHb2NrE0+u0/tpRUzxjDC9fvoRIJKrxOB9Np6bixmQmJiZqFa9Ro0Zqv1lrG6+p2ArGxsZ1nmNDeT4A9fPTZDs0+XwAmqlhbeMpxw9rB0DvxfpoB71OP6wdgG68TtUZjABoonCVZs+erbF4TcXWlibboe85avL50ORxKcf6ia2thlIXXcyRXqcf1o7a0IW68Jg6M2/0QGFhIUxMTPDixYta92J1hb7nqO/5AZSjvtD3HPU9P4By1FUfzUiNUCjEypUrIRQKtd0UjdH3HPU9P4By1Bf6nqO+5wdQjrrqoxmpIYQQQoh++2hGagghhBCi36hTQwghhBC9QJ0aQgghhOgFnerUvHnzBomJiQgICICpqSkiIiKUHv/9998xYMAAiEQiiMViBAQEoKSkRCkmJiYGn332GSQSCWxsbPDVV1/hwYMHSjGJiYno27cvrK2t0a5dO+zcuVPTqXHqK0djY2OIRCJIJBLuZ9GiRZpOr07yO3jwILp06QJLS0u0b98e27dvVzmPrtdQnRy1VUMACA8PR6dOnSAWi+Hg4IBt27YpPS6TybBkyRLY2dlBJBLBy8sLmZmZSjGZmZkYO3Ys2rRpA7FYjAULFkAmkynFaLOO9ZWjrtcxJycHkZGR6NevH6RSaaXn0VYd6ys/Xa9hRkYGxo4dCysrK0gkEowYMQJpaWlKMdp8L9YK0yHff/8969GjB1u2bBkzMzNju3fv5h57+vQpa9myJQsLC2Nv3rxhBQUFzMvLi/n5+XExV65cYUKhkP3rX/9ijDFWUlLC5s+fzyQSCXv9+jVjjLG7d+8yIyMjFhUVxRhj7Pbt26xVq1bsn//8p97kWFBQwHg8HispKamXnN72ofmdPn2aGRsbs4sXLzLGGLt37x5r3749i4iI4GJ0vYbq5KjNGv74449MIpGwW7duMcbKn29LS0u2d+9eLmbKlCmsX79+LD8/n8nlcrZgwQLWuXNnVlZWxhhjTCaTsQ4dOrCAgAAml8tZfn4+69u3L/vzn//MHUObdayvHHW9jowx1rVrVzZu3Dg2YcIEZmNjo3IebdWxvvLT9RqWlpYye3t7tmjRIlZaWsrKyspYYGAg69ChA5PL5dxxtfk3tTZ0qlPzNhsbG6UPi127djEHBwelmLy8PMbn89mTJ08YY4xt2LCBOTk5KcUUFBQwAOzKlSuMMcamTp3Khg0bphTz7bffquxXHzSV461bt5iFhYVmG6+G98lvwoQJbObMmUox0dHRrFOnTtzvul5DdXLUZg1nzZrF9u3bp7QtICCAjRw5kjHGWHp6OjMwMGBJSUnc4zKZjLVs2ZIdPnyYMcbYnj17mKmpKZPJZFxMUlISEwgELCcnhzGm3TrWV466Xse37d69u9IPfW3Vsb7y0/Ua3rhxg/Xv35+9efOGiyksLGQA2PXr1xljDetvak106uun6rx48YJbCqGCUCgEYwwJCQkAAGdnZ9y/fx/37t3jYo4dOwYLCwu0b98eABAbGwtPT0+l43h6euLatWt4+vSphrOoXl3lmJGRAYlEUn8NV5M6+VUW07RpU9y6dQsFBQUAdL+G6uSozRqGhYVh/PjxSttu3rzJ3bzr3LlzaNWqFZydnbnHBQIBvvjiCxw/fhxAeY0GDRoEgUDAxTg7O6Nly5aIiYnhYrRVx/rKUdfrqA5t1bG+8tP1Gnbu3Blnz54Fj8dTOgYAGBkZAWjYf1PfpTedmhEjRuDhw4fYsmUL5HI5nj9/Dn9/fzRr1gxPnjwBALi7u2Pr1q0YMmQIpkyZgmHDhiEqKgpxcXFo3rw5ACArK0tl0SyxWAwAKt9D1re6yjEjIwNCoRCzZ8+Gg4MDOnbsiKVLl+L169faTE+t/MaPH489e/YgNjYWjDEkJydj3bp1AMDF6HoN1cmxodRQLpdj7ty5SEhIQGBgIIDKn3+gvAYVz//7xmijjprMUdfrqI6GUEdN5qdvNbxy5QrGjBkDPz8/bg5RQ6ihuvSmU2Nra4vz58/jxIkTsLW1xcCBAzFq1CjY2NigcePydTsVCgUePXqE1q1bo0ePHujevTtu3LiB8+fPc8fh8/kq/0p+uwerTXWVo0wmw8uXL+Hj44Pbt2/j1KlTSEhIwNSpU7WVGgD18vP29sb27duxePFi2NjYYO7cuVi9ejUAcDG6XkN1cmwINUxPT0ffvn0RExOD+Ph4dOrUCUDlzz+gXIP3janvOmo6R12vozq0XUdN56dPNQwNDUXfvn3h5+eH77//ntuu7RrWija/+/oQ785VqIxcLmdNmzZlsbGxjDHG1q1bx1xcXFhpaSkXk5qayoyMjNi5c+cYY4w5ODiwXbt2KR3njz/+YAC4OQ/1RVM5ViYxMZEZGBiwoqKiOmm7Ot4nv8rEx8czPp/PTdTT9RpW5t0cK1OfNUxKSmLm5uZs4cKFKm06cOAAE4vFKvv4+Phwc4VmzpzJJkyYoBIjkUjYgQMHGGPar2N95FgZXarj26qac6LNOtZHfpXRtRoqFAo2ZcoU1q5dO5aYmKgSr+33Ym3ozUgNAJXhvpMnT0IoFKJXr14AgPj4ePTu3Rt8Pp+LadOmDezs7HDp0iUAwODBg/Hrr78qHef06dPo1q0bWrVqpeEMalYXOQLllx2/TaFQANB+77um/CqL+fe//w03Nzdu/RJdr2FlMe/mCGivhunp6RgyZAjCwsLw7bffqqwbM2DAADx79gw3btxQatvZs2fh4eEBoLxGp06dQllZGRdz7949PHv2DG5ublyMtupYXzkCul1HdWirjvWVH6D7NfzLX/6C+/fvIykpCZ9++qnKeRr631Ql2u5Vva93/wV88+ZNZmFhwX777TfGWHkvsl27diwsLIyL2bhxI2vdujUXU1ZWxnbu3Mn4fD43OzwlJYUZGxtzM8OTk5OZWCxm+/fvr6fM/kdTOW7YsIENHDiQZWZmMsYYy8rKYr1792YTJ06sp8zKvU9+O3fuZN27d+f+dXDixAn2ySefcFd2Mab7NVQnR23W0MPDgwUFBVUbM336dObm5sZevHjBysrK2OLFi5mjoyN3iahcLmeOjo5syZIlrKysjBUUFDA3Nzc2Y8YM7hjarGN95ajrdXxbVSMZ2qpjfeWn6zVMTExkZmZmLDc3t8pjNKS/qTXRm04NY+Uvug4dOjCRSMTs7e3Zzp07lR5XKBRsy5YtrHPnzkwkErFWrVoxd3d3FhMToxR34cIF5uLiwkQiEbOzs2M7duzQdDqV0lSOxcXFbNmyZczW1pZZWloykUjE5s2bx93Hpr68T34ymYwtXryYWVtbM5FIxHr27Mni4uJUjq3LNVQnR23WEACzsLBgYrFY5adCxf2RxGIxa926NfPy8mKPHz9WOs7jx4+Zl5cXs7S0ZGKxmM2fP19l+FxbdayvHPWhjhWq+3pGG3Wsr/x0vYZBQUGsSZMmlR5j06ZNXFxD+ZtaE1qlmxBCCCF6Qa/m1BBCCCHk40WdGkIIIYToBerUEEIIIUQvUKeGEEIIIXqBOjWEEEII0QvUqSGEEEKIXqBODSGEEEL0AnVqCCGEEKIXqFNDCCGEEL1AnRpCCCGE6AXq1BBCCCFEL1CnhhBCCCF64f8B9dbfC5YkJvgAAAAASUVORK5CYII=",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"from myst_nb import glue\n",
"t_shift = 6\n",
@@ -814,7 +1080,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 10,
"metadata": {
"editable": true,
"hidden": true,
@@ -823,7 +1089,25 @@
},
"tags": []
},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "GDPと6期先のデフレータの乖離率の相関係数:0.433\n"
+ ]
+ },
+ {
+ "data": {
+ "image/png": "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",
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
"source": [
"t_shift = 6\n",
"tmp = df.copy()\n",
@@ -880,7 +1164,10 @@
"source": [
"定量分析を進めるためには,ADASモデルのパラメータに値を設定する必要がある。\n",
"ここでは景気循環のデータに基づき,次の4つのパラメーターの値を決めていく\n",
- "* $a$,$c$,$\\sigma_u$,$\\sigma_v$ (標準偏差)\n",
+ "* $a$ (AS曲線の傾き)\n",
+ "* $c$ (AD曲線の傾き)\n",
+ "* $\\sigma_v$ (供給ショックの標準偏差)\n",
+ "* $\\sigma_u$ (需要ショックの標準偏差)\n",
"\n",
"パラメーターの値の決め方には主に次の2つある。\n",
"1. データに基づいて計量経済学的な手法などを用いて推定した結果を使う。\n",
@@ -960,13 +1247,21 @@
"$$ (eq:18-ad_small2)"
]
},
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "カリブレーションでは,パラメータ`a`と`c`を推定する必要があるが,上の式を回帰式として使うには同時性の問題がある。\n",
+ "式[](eq:18-as_small2)を$\\Delta p_{t}=ay_{t}+v_{t}$,$\\Delta p_{t}\\equiv p_{t}-p_{t-1}$に変形して回帰式として解釈し,$a$を推定することが考えられる。しかし,供給ショック$v_t$は$p_t$に影響を与え,式[](eq:18-ad_small2)を通して$y_t$にも影響を及ぼすことが考えられる。即ち,$v_t$と$y_t$が相関し,このままでは$a$の推定値は不偏性も一致性も失うことになる。また,式[](eq:18-ad_small2)をそのまま回帰式として解釈し,$c$を推定しても類似の問題が考えられる。"
+ ]
+ },
{
"cell_type": "markdown",
"metadata": {
"hidden": true
},
"source": [
- "モデルの安定性を考えるために,式[](eq:18-ad_small)を式[](eq:18-as_small)に代入すると,次のように整理することができる。\n",
+ "この点を回避するために,式[](eq:18-ad_small)を式[](eq:18-as_small)に代入し次のように整理する。\n",
"\n",
"$$\\begin{aligned}\n",
"p_{t}\n",
@@ -988,17 +1283,17 @@
"hidden": true
},
"source": [
- "また,式[](eq:18-eq-ppp)を[](eq:18-ad_small2)に代入し,$h$の定義を利用し整理すると次式となる。\n",
+ "また,式[](eq:18-eq-ppp)を[](eq:18-ad_small2)に代入し,`k`の定義を利用し整理すると次式となる。\n",
"\n",
"$$\n",
"\\begin{aligned}\n",
"y_t\n",
"&=-c\\left[hp_{t-1}+h\\left(a u_{t}+v_{t}\\right)\\right]+u_t\\\\\n",
"&=-ch p_{t-1}-ch\\left(a u_{t}+v_{t}\\right)+u_t\\\\\n",
- "&=-ch p_{t-1}+(1-ach)u_t-chv_t\\\\\n",
- "&=-ch p_{t-1}+\\left(1-\\dfrac{ac}{1+ac}\\right)u_t-chv_t\\\\\n",
- "&=-ch p_{t-1}+\\dfrac{1}{1+ac}u_t-chv_t\\\\\n",
- "&=-ch p_{t-1}+hu_t-chv_t\n",
+ "&=-ch p_{t-1}+(1-ach)u_t-ckv_t\\\\\n",
+ "&=-ch p_{t-1}+\\left(1-\\dfrac{ac}{1+ac}\\right)u_t-ckv_t\\\\\n",
+ "&=-ch p_{t-1}+\\dfrac{1}{1+ac}u_t-ckv_t\\\\\n",
+ "&=-ch p_{t-1}+hu_t-ckv_t\n",
"\\end{aligned}\n",
"$$"
]
@@ -1009,7 +1304,7 @@
"hidden": true
},
"source": [
- "この結果を利用して,以下ではADASモデルを次の2つの式で表すことにする。"
+ "この結果を利用すると,ADASモデルを次の2つの式で表すことができる。"
]
},
{
@@ -1019,7 +1314,9 @@
},
"source": [
"$$\n",
- "p_{t}=hp_{t-1}+h\\left(a u_{t}+v_{t}\\right)\n",
+ "p_{t}=hp_{t-1}+e_{pt},\n",
+ "\\qquad\\quad\n",
+ "e_{pt}\\equiv k\\left(a u_{t}+v_{t}\\right)\n",
"$$ (eq:18-eq-p)"
]
},
@@ -1030,7 +1327,9 @@
},
"source": [
"$$\n",
- "y_t = -chp_{t-1} + h(u_t-cv_t)\n",
+ "y_t = -chp_{t-1} + e_{yt}\n",
+ "\\qquad\\quad\n",
+ "e_{yt}\\equiv k(u_t-cv_t)\n",
"$$ (eq:18-eq-y)"
]
},
@@ -1047,50 +1346,155 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "次に,パラメータ$a$と$c$の推定値を求めるが,次の問題がある。式[](eq:18-as_small2)を$\\Delta p_{t}=ay_{t}+v_{t}$,$\\Delta p_{t}\\equiv p_{t}-p_{t-1}$に変形して回帰式として解釈し,$a$を推定することが考えられる。しかし,供給ショック$v_t$は$p_t$に影響を与え,それを通して$y_t$にも影響を及ぼすことが考えられる。即ち,$v_t$と$y_t$が相関し,このままでは$a$の推定値は不偏性も一致性も失うことになる。また,式[](eq:18-ad_small2)をそのまま回帰式として解釈し,$c$を推定しても同様の問題が考えられる。"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "従って,このような問題を避けるために,式[](eq:18-eq-p)を回帰式と捉えて,まずは$h$の推定値を計算する。式[](eq:18-eq-p)は$p_t$の差分方程式となっており,次のような自己回帰モデルとなっている。\n",
- "\n",
- "$$\n",
- "p_{t}=hp_{t-1} + e_{pt}\n",
- "$$ (eq:18-regression-h)\n",
- "\n",
- "* $e_{pt}\\equiv h\\left(a u_{t}+v_{t}\\right)$\n",
- "\n",
- "説明変数である$p_{t-1}$は誤差項$e_{pt}$とは期間がズレているため,仮定に基づくと$\\text{E}(e_{pt}|p_{t-1})=0$が満たされ,推定値は一致性を満たすことになる(不偏性は満たさない)。\n",
- "\n",
+ "式[](eq:18-eq-p)を回帰式と捉えて,まずは$h$の推定値を計算する。\n",
"まず,`df`のメソッド`.shift()`を使って`deflator_cycle`を1期ずらした列を`deflator_cycle_lag`として`df`に追加しよう。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 11,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "\n",
+ "\n",
+ "
\n",
+ " \n",
+ " \n",
+ " | \n",
+ " deflator | \n",
+ " gdp_cycle | \n",
+ " deflator_cycle | \n",
+ " deflator_cycle_lag | \n",
+ "
\n",
+ " \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ " | \n",
+ "
\n",
+ " \n",
+ " \n",
+ " \n",
+ " 1980-01-01 | \n",
+ " 91.2 | \n",
+ " 0.003269 | \n",
+ " -0.025798 | \n",
+ " NaN | \n",
+ "
\n",
+ " \n",
+ " 1980-04-01 | \n",
+ " 93.4 | \n",
+ " -0.010845 | \n",
+ " -0.006555 | \n",
+ " -0.025798 | \n",
+ "
\n",
+ " \n",
+ " 1980-07-01 | \n",
+ " 94.4 | \n",
+ " 0.000457 | \n",
+ " -0.000482 | \n",
+ " -0.006555 | \n",
+ "
\n",
+ " \n",
+ " 1980-10-01 | \n",
+ " 95.4 | \n",
+ " 0.010468 | \n",
+ " 0.005515 | \n",
+ " -0.000482 | \n",
+ "
\n",
+ " \n",
+ " 1981-01-01 | \n",
+ " 95.6 | \n",
+ " 0.009441 | \n",
+ " 0.003126 | \n",
+ " 0.005515 | \n",
+ "
\n",
+ " \n",
+ "
\n",
+ "
"
+ ],
+ "text/plain": [
+ " deflator gdp_cycle deflator_cycle deflator_cycle_lag\n",
+ " \n",
+ "1980-01-01 91.2 0.003269 -0.025798 NaN\n",
+ "1980-04-01 93.4 -0.010845 -0.006555 -0.025798\n",
+ "1980-07-01 94.4 0.000457 -0.000482 -0.006555\n",
+ "1980-10-01 95.4 0.010468 0.005515 -0.000482\n",
+ "1981-01-01 95.6 0.009441 0.003126 0.005515"
+ ]
+ },
+ "execution_count": 11,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"df['deflator_cycle_lag'] = df['deflator_cycle'].shift()\n",
- "df.iloc[:5,-5:]"
+ "df.iloc[:5,-4:]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`deflator_cycle_lag`には1期前の`deflator_cycle`の値が並んでいるのが確認できる。\n",
+ "`deflator_cycle_lag`には1期前の`deflator_cycle`の値が,`d_deflator_cycle`には前期との差分が並んでいるのが確認できる。\n",
"\n",
"まず回帰分析をおこない,その結果を表示する。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 12,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " OLS Regression Results \n",
+ "==============================================================================\n",
+ "Dep. Variable: deflator_cycle R-squared: 0.689\n",
+ "Model: OLS Adj. R-squared: 0.688\n",
+ "Method: Least Squares F-statistic: 384.1\n",
+ "Date: Sat, 11 Jan 2025 Prob (F-statistic): 8.56e-46\n",
+ "Time: 15:36:45 Log-Likelihood: 716.66\n",
+ "No. Observations: 175 AIC: -1429.\n",
+ "Df Residuals: 173 BIC: -1423.\n",
+ "Df Model: 1 \n",
+ "Covariance Type: nonrobust \n",
+ "======================================================================================\n",
+ " coef std err t P>|t| [0.025 0.975]\n",
+ "--------------------------------------------------------------------------------------\n",
+ "Intercept 0.0002 0.000 0.795 0.428 -0.000 0.001\n",
+ "deflator_cycle_lag 0.8194 0.042 19.597 0.000 0.737 0.902\n",
+ "==============================================================================\n",
+ "Omnibus: 48.339 Durbin-Watson: 1.616\n",
+ "Prob(Omnibus): 0.000 Jarque-Bera (JB): 109.483\n",
+ "Skew: 1.218 Prob(JB): 1.68e-24\n",
+ "Kurtosis: 6.014 Cond. No. 136.\n",
+ "==============================================================================\n",
+ "\n",
+ "Notes:\n",
+ "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
+ ]
+ }
+ ],
"source": [
"res_h = smf.ols('deflator_cycle ~ deflator_cycle_lag', data=df).fit()\n",
"print(res_h.summary())"
@@ -1105,7 +1509,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 13,
"metadata": {
"editable": true,
"slideshow": {
@@ -1115,76 +1519,118 @@
"remove-cell"
]
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "application/papermill.record/text/plain": "1.616"
+ },
+ "metadata": {
+ "scrapbook": {
+ "mime_prefix": "application/papermill.record/",
+ "name": "durbin_watson_h"
+ }
+ },
+ "output_type": "display_data"
+ }
+ ],
"source": [
"from statsmodels.stats import stattools\n",
"res_hg = smf.ols('deflator_cycle ~ deflator_cycle_lag', data=df).fit()\n",
"dw_h = stattools.durbin_watson(res_hg.resid)\n",
- "glue(\"durbin_watson_h\", round(dw_h.item(),3), display=False)"
+ "glue(\"durbin_watson_h\", round(dw_h.item(),3), display=False)\n",
+ "# print(dw_h)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "残差の自己相関を考えてみよう。ダービン・ワトソン検定統計量は{glue:}`durbin_watson_h`であり,「不明」の領域にある。しかし,[付録B](sec:18-1-appendix_B)ではBreusch-Godfrey検定とLjung-Box検定をおこなっているが,帰無仮説である「残差の自己相関なし」は棄却されてしまう。この結果により,説明変数`deflator_cycle_lag`と誤差項の相関の疑いがあり,`h`の推定値は不偏性も一致性も満たさないことになってしまう。この問題を解決するために,一般化最小二乗法の一つであるコクラン=オーカット推定法(Cochrane-Orcutt推定方)を使い推定をおこなうことにする。\n",
+ "\n",
+ "コードは簡単で,`statsmodels`の`.ols()`関数の代わりに`.glsar()`関数を使う(`glsar`は**G**eneralized **L**inear **S**quares with **A**utoreg**r**essive Errorsの略)。また,未知の自己相関係数を計算する反復推定をおこなうために,`.fit()`の代わりに次の引数を使い`iterative_fit()`を使用する。\n",
+ "* `itermax=10`:反復計算を最大で`10`回おこなう。\n",
+ "* `cov_type='HC1'`:不均一分散の可能性があるため不均一分散頑健推定を使う。\n",
+ " * 不均一分散頑健推定については[Pythonで学ぶ入門計量経済学ー不均一分散](https://py4etrics.github.io/14_Hetero.html)を参照しよう。"
]
},
{
"cell_type": "code",
- "execution_count": null,
- "metadata": {
- "tags": [
- "remove-cell"
- ]
- },
- "outputs": [],
+ "execution_count": 40,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " GLSAR Regression Results \n",
+ "==============================================================================\n",
+ "Dep. Variable: deflator_cycle R-squared: 0.641\n",
+ "Model: GLSAR Adj. R-squared: 0.639\n",
+ "Method: Least Squares F-statistic: 175.0\n",
+ "Date: Sat, 11 Jan 2025 Prob (F-statistic): 5.17e-28\n",
+ "Time: 16:23:23 Log-Likelihood: 720.75\n",
+ "No. Observations: 174 AIC: -1438.\n",
+ "Df Residuals: 172 BIC: -1431.\n",
+ "Df Model: 1 \n",
+ "Covariance Type: HC1 \n",
+ "======================================================================================\n",
+ " coef std err z P>|z| [0.025 0.975]\n",
+ "--------------------------------------------------------------------------------------\n",
+ "Intercept 0.0002 0.000 0.448 0.654 -0.001 0.001\n",
+ "deflator_cycle_lag 0.8188 0.062 13.230 0.000 0.697 0.940\n",
+ "==============================================================================\n",
+ "Omnibus: 53.697 Durbin-Watson: 2.037\n",
+ "Prob(Omnibus): 0.000 Jarque-Bera (JB): 151.719\n",
+ "Skew: 1.257 Prob(JB): 1.13e-33\n",
+ "Kurtosis: 6.822 Cond. No. 137.\n",
+ "==============================================================================\n",
+ "\n",
+ "Notes:\n",
+ "[1] Standard Errors are heteroscedasticity robust (HC1)\n"
+ ]
+ }
+ ],
+ "source": [
+ "res_h_glsar = smf.glsar('deflator_cycle ~ deflator_cycle_lag',\n",
+ " data=df).iterative_fit(itermax=10,\n",
+ " cov_type='HC1')\n",
+ "print(res_h_glsar.summary())"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
"source": [
- "from statsmodels.stats.api import het_breuschpagan, het_white\n",
- "breuschpagan_test_h = het_breuschpagan(res_hg.resid, res_hg.model.exog)[1]\n",
- "white_test_h = het_white(res_hg.resid, res_hg.model.exog)[1]\n",
- "glue(\"breuschpagan_h\", round(breuschpagan_test_h.item(), 6), display=False)\n",
- "glue(\"white_h\", round(white_test_h.item(), 7), display=False)"
+ "`h`の推定値は大きく変わらないが,残差の系列相関はそれ程大きくないためだ。統計的優位性も高く,コクラン=オーカット推定法による式の変形後の残差の系列相関も排除できる。またこの計算では,不均一分散の可能性も排除できないため,不均一分散頑健推定も使っており,$t$検定は有効と考えることができる。ここで計算した値を`h`の推定値として採用し変数`hhat`に割り当てることにする。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 42,
"metadata": {
"tags": [
- "remove-cell"
+ "hide-input"
]
},
"outputs": [],
"source": [
- "res_hg_hac = res_hg.get_robustcov_results(cov_type='HAC', maxlags=1, use_t=True)\n",
- "pval_h_lag = res_hg_hac.pvalues[1]\n",
- "glue(\"pval_h\", round(pval_h_lag.item(), 10), display=False)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "* 残差の自己相関\n",
- " * 式[](eq:18-regression-h)は自己回帰モデルとなるため,ダービン・ワトソン検定統計量($d$検定){glue:}`durbin_watson_h`は無効となる。その代わりに,[付録A](sec:18-1-appendix_A)ではBreusch-Godfrey検定とLjung-Box検定をおこなっているが,残差の自己相関を棄却できない。従って,`t`検定は有効ではない。\n",
- "* 残差の均一分散([付録A](sec:18-1-appendix_A)を参照)\n",
- " * ブルーシュペーガン検定の$p$値は{glue:}`breuschpagan_h`であり、とホワイト検定の値は{glue:}`white_h`となり,帰無仮説(均一分散)は「通常」の優位性水準では棄却される。\n",
- "* 不均一分散自己相関頑健推定(`HAC`)\n",
- " * この手法を使うと推定値`h`の`t`検定は有効になる。その場合の`deflator_cycle_lag`の$p$値は{glue:}`pval_h`であり([付録A](sec:18-1-appendix_A)を参照)、推定値の統計的優位性は高いことが確認できる。\n",
- "* 定数項なしで推定しても結果は殆ど変わらない。(試してみよう!)"
+ "hhat = res_h_glsar.params['deflator_cycle_lag']"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "式[](eq:18-eq-p)は`AR(1)`であるため,$\\hat{h}$は$p_t$の自己相関係数でもあり,非常に高い持続性が特徴となっている。\n",
- "\n",
- "以下に続く分析のために,次の変数を作成しておこう。"
+ "また以下の分析では残差も必要になるが,コクラン=オーカット推定法で計算した$\\hat{e}_{pt}$を変数`ep`に割り当てることにする。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 44,
"metadata": {},
"outputs": [],
"source": [
- "hhat = res_h.params.iloc[1] # hの推定値\n",
- "ep = res_h.resid # 推定式(3)の残差"
+ "ep = res_h_glsar.resid"
]
},
{
@@ -1200,7 +1646,6 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "上で説明した問題があるため,式[](eq:18-ad_small2)を回帰式として$c$を推定することはできない。\n",
"ここでは`hhat`を利用して`c`の推定値を計算するために,式[](eq:18-eq-y)を回帰式として使う。\n",
"式[](eq:18-eq-y)を再掲しよう。\n",
"\n",
@@ -1211,14 +1656,46 @@
"* $d\\equiv -ch$\n",
"* $e_{yt}\\equiv hu_t-chv_t$\n",
"\n",
- "説明変数である$p_{t-1}$は誤差項$e_{yt}$とは期間がズレているため,仮定に基づくと$\\text{E}(e_{yt}|p_{t-1})=0$となり,推定値は一致性を満たすことになる(不偏性は満たさない)。"
+ "説明変数である$p_{t-1}$は誤差項$e_{yt}$とは期間がズレているため,残差に系列相関がなければ推定値は一致性を満たすことになる(不偏性は満たさない)。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 18,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " OLS Regression Results \n",
+ "==============================================================================\n",
+ "Dep. Variable: gdp_cycle R-squared: 0.041\n",
+ "Model: OLS Adj. R-squared: 0.035\n",
+ "Method: Least Squares F-statistic: 7.398\n",
+ "Date: Sat, 11 Jan 2025 Prob (F-statistic): 0.00719\n",
+ "Time: 15:39:21 Log-Likelihood: 492.28\n",
+ "No. Observations: 175 AIC: -980.6\n",
+ "Df Residuals: 173 BIC: -974.2\n",
+ "Df Model: 1 \n",
+ "Covariance Type: nonrobust \n",
+ "======================================================================================\n",
+ " coef std err t P>|t| [0.025 0.975]\n",
+ "--------------------------------------------------------------------------------------\n",
+ "Intercept -6.676e-05 0.001 -0.060 0.952 -0.002 0.002\n",
+ "deflator_cycle_lag -0.4099 0.151 -2.720 0.007 -0.707 -0.112\n",
+ "==============================================================================\n",
+ "Omnibus: 58.282 Durbin-Watson: 0.633\n",
+ "Prob(Omnibus): 0.000 Jarque-Bera (JB): 236.393\n",
+ "Skew: -1.212 Prob(JB): 4.66e-52\n",
+ "Kurtosis: 8.152 Cond. No. 136.\n",
+ "==============================================================================\n",
+ "\n",
+ "Notes:\n",
+ "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
+ ]
+ }
+ ],
"source": [
"res_d = smf.ols('gdp_cycle ~ deflator_cycle_lag', data=df).fit()\n",
"print(res_d.summary())"
@@ -1251,71 +1728,85 @@
]
},
{
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "tags": [
- "remove-cell"
- ]
- },
- "outputs": [],
+ "cell_type": "markdown",
+ "metadata": {},
"source": [
- "from statsmodels.stats.api import het_breuschpagan, het_white\n",
- "breuschpagan_test_d = het_breuschpagan(res_dg.resid, res_dg.model.exog)[1]\n",
- "white_test_d = het_white(res_dg.resid, res_dg.model.exog)[1]\n",
- "glue(\"breuschpagan_d\", round(breuschpagan_test_d.item(), 6), display=True)\n",
- "glue(\"white_d\", round(white_test_d.item(), 7), display=True)"
+ "$d$の推定値である$\\hat{d}$は負であり,モデルに沿った結果である。\n",
+ "しかし,`Durbin-Watson`検定統計量は{glue:}`durbin_watson_d`であり残差の正の系列相関が強く疑われる。\n",
+ "従って,`deflator_cycle_lag`と残差の相関の可能性があり,推定値の一致性が失われている可能性が高い。\n",
+ "この問題に対処するために,`h`の推定値を計算した時と同じように,コクラン=オーカット推定法を使い推定をおこなう。\n",
+ "結果は次の通りである。"
]
},
{
"cell_type": "code",
- "execution_count": null,
- "metadata": {
- "tags": [
- "remove-cell"
- ]
- },
- "outputs": [],
- "source": [
- "res_d_hac = smf.ols('gdp_cycle ~ deflator_cycle_lag', data=df).fit(cov_type='HAC', cov_kwds={'maxlags':1}, use_t=True)\n",
- "pval_d_hac = res_d_hac.pvalues['deflator_cycle_lag']\n",
- "glue(\"pval_d_hac\", round(pval_d_hac.item(), 10), display=True)"
- ]
- },
- {
- "cell_type": "markdown",
+ "execution_count": 45,
"metadata": {},
- "source": [
- "* `Durbin-Watson`検定量は{glue:}`durbin_watson_d`であり残差の正の系列相関が疑われる。\n",
- " * $t$検定は無効の可能性を拭えない。\n",
- "* 残差の均一分散([付録B](sec:18-1-appendix_B)を参照)\n",
- " * ブルーシュペーガン検定の$p$値は{glue:}`breuschpagan_d`であり、帰無仮説(均一分散)は有意水準`5`%で棄却される。一方、ホワイト検定の$p$値は{glue:}`white_d`となり、帰無仮説(均一分散)は有意水準`5`%で棄却できないが、`10`%では棄却される。従って、不均一分散の可能性も残っている。従って,この理由からも$t$検定は無効になるかも知れない。\n",
- "* 不均一分散自己相関頑健推定(`HAC`)を使うと推定値`h`の`t`検定は有効になる。その場合の`deflator_cycle_lag`の$p$値は{glue:}`pval_d_hac`であり、推定値の統計的優位性は高いことが確認できる([付録B](sec:18-1-appendix_B)を参照)。この結果に基づき`d`の推定値を採用する。\n",
- "* 定数項なしで推定しても結果は殆ど変わらない。"
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " GLSAR Regression Results \n",
+ "==============================================================================\n",
+ "Dep. Variable: gdp_cycle R-squared: 0.007\n",
+ "Model: GLSAR Adj. R-squared: 0.001\n",
+ "Method: Least Squares F-statistic: 0.5500\n",
+ "Date: Sat, 11 Jan 2025 Prob (F-statistic): 0.459\n",
+ "Time: 16:33:46 Log-Likelihood: 544.68\n",
+ "No. Observations: 174 AIC: -1085.\n",
+ "Df Residuals: 172 BIC: -1079.\n",
+ "Df Model: 1 \n",
+ "Covariance Type: HC1 \n",
+ "======================================================================================\n",
+ " coef std err z P>|z| [0.025 0.975]\n",
+ "--------------------------------------------------------------------------------------\n",
+ "Intercept 0.0004 0.003 0.147 0.883 -0.005 0.005\n",
+ "deflator_cycle_lag -0.2202 0.297 -0.742 0.458 -0.802 0.362\n",
+ "==============================================================================\n",
+ "Omnibus: 151.661 Durbin-Watson: 2.052\n",
+ "Prob(Omnibus): 0.000 Jarque-Bera (JB): 3130.415\n",
+ "Skew: -3.044 Prob(JB): 0.00\n",
+ "Kurtosis: 22.867 Cond. No. 75.8\n",
+ "==============================================================================\n",
+ "\n",
+ "Notes:\n",
+ "[1] Standard Errors are heteroscedasticity robust (HC1)\n"
+ ]
+ }
+ ],
+ "source": [
+ "res_d_glsar = smf.glsar('gdp_cycle ~ deflator_cycle_lag', data=df).iterative_fit(10, cov_type='HC1')\n",
+ "print(res_d_glsar.summary())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "$d$の推定値である$\\hat{d}$は負であり,モデルに沿った結果である。\n",
- "推定値を`dhat`に割り当てることにする。"
+ "`d`の推定値は絶対値で大きく減少しており,系列相関の影響が大きいことを示している。\n",
+ "不均一分散の可能性も排除できないため,不均一分散頑健推定を使っているが,残念ながら統計的優位性は低い。\n",
+ "この値を`d`の推定値として変数`dhat`に使うことにする。"
]
},
{
"cell_type": "code",
- "execution_count": null,
- "metadata": {},
+ "execution_count": 46,
+ "metadata": {
+ "tags": [
+ "hide-input"
+ ]
+ },
"outputs": [],
"source": [
- "dhat = res_d.params.iloc[1]"
+ "dhat = res_d_glsar.params['deflator_cycle_lag']"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "ここで重要なのは,$-d\\equiv ch$となり,$c$の推定値を次式で計算することができる。\n",
+ "`d`の推定値が確定した訳だが,`d`$-d\\equiv ch$で与えられ,$c$の推定値を次式で計算することができる。\n",
"\n",
"$$\n",
"\\hat{c} = -\\frac{\\hat{d}}{\\hat{h}}\n",
@@ -1326,9 +1817,20 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 48,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "0.2689802972319807"
+ ]
+ },
+ "execution_count": 48,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"chat = -dhat / hhat\n",
"chat"
@@ -1338,12 +1840,13 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "これが「最初のスイッチの調整」である。また後に続く計算のために推定結果の残差を割りてて次の変数も作成しておく。"
+ "これが「最初のスイッチの調整」である。\n",
+ "また以下の分析では残差も必要になるため,コクラン=オーカット推定法で計算した$\\hat{e}_{yt}$を変数`ep`に割り当てることにする。"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 49,
"metadata": {},
"outputs": [],
"source": [
@@ -1379,11 +1882,22 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 50,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "0.8227478597459131"
+ ]
+ },
+ "execution_count": 50,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"ahat = (1/hhat - 1) / chat\n",
"ahat"
@@ -1440,7 +1954,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 51,
"metadata": {},
"outputs": [],
"source": [
@@ -1457,9 +1971,20 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 52,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "0.012546750971860659"
+ ]
+ },
+ "execution_count": 52,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"v_std = vt.std()\n",
"v_std"
@@ -1494,11 +2019,22 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 53,
"metadata": {
"hidden": true
},
- "outputs": [],
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "0.014633934415485182"
+ ]
+ },
+ "execution_count": 53,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
"ut = ey + chat*ep\n",
"u_std = ut.std()\n",
@@ -1511,10 +2047,7 @@
"hidden": true
},
"source": [
- "この値を$\\sigma_u$に使うが,これが「最後のスイッチの調整」となる。\n",
- "\n",
- "`v_std`と`u_std`を比べると,後者の値がより大きい。\n",
- "ADASモデルの枠組みで考えると,需要ショックの変動幅が大きいことを示している。"
+ "この値を$\\sigma_u$に使うが,これが「最後のスイッチの調整」となる。"
]
},
{
@@ -1528,127 +2061,116 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 60,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "aの値:0.823\n",
+ "cの値:0.269\n",
+ "vの標準偏差:0.012547\n",
+ "uの標準偏差:0.014634\n"
+ ]
+ }
+ ],
"source": [
"print(f'aの値:{ahat:.3f}')\n",
"print(f'cの値:{chat:.3f}')\n",
- "print(f'uの標準偏差:{u_std:.6f}')\n",
- "print(f'vの標準偏差:{v_std:.6f}')"
+ "print(f'vの標準偏差:{v_std:.6f}')\n",
+ "print(f'uの標準偏差:{u_std:.6f}')"
]
},
{
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
+ "cell_type": "code",
+ "execution_count": 62,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "aの値:0.82274785974591313487\n",
+ "cの値:0.26898029723198069174\n",
+ "vの標準偏差:0.01254675097186065853\n",
+ "uの標準偏差:0.01463393441548518220\n"
+ ]
+ }
+ ],
"source": [
- "需要ショック($u_t$)の標準偏差が比較的に大きく,供給ショック($v_t$)の2倍以上となっている。この結果は,以下で考察する定量的な問に関する結果を理解する鍵となる。\n",
- "\n",
- "次章では,これらの値を使い定量分析を更に進めていく。"
+ "print(f'aの値:{ahat:.20f}')\n",
+ "print(f'cの値:{chat:.20f}')\n",
+ "print(f'vの標準偏差:{v_std:.20f}')\n",
+ "print(f'uの標準偏差:{u_std:.20f}')"
]
},
{
"cell_type": "markdown",
"metadata": {
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
- "## 付録"
+ "`v_std`と`u_std`を比べると,後者の値が若干大きい様である。\n",
+ "比率を計算してみよう。"
]
},
{
- "cell_type": "markdown",
- "metadata": {
- "jp-MarkdownHeadingCollapsed": true
- },
+ "cell_type": "code",
+ "execution_count": 58,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "1.166352504190573"
+ ]
+ },
+ "execution_count": 58,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
- "(sec:18-1-appendix_A)=\n",
- "### 付録A"
+ "u_std / v_std"
]
},
{
"cell_type": "markdown",
"metadata": {
- "jp-MarkdownHeadingCollapsed": true
+ "hidden": true
},
"source": [
- "#### 残差の自己相関の検定"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "式[](eq:18-regression-h)は自己回帰モデルとなるため,(通常の)ダービン・ワトソン検定($d$検定)は無効となる。その代わりに,Breusch-Godfrey検定を使う。\n",
- "\n",
- "<帰無仮説>\n",
- "\n",
- "$$\n",
- "e_{pt}=\\rho_{p}e_{pt-1}+e_{pt}\n",
- "$$\n",
- "\n",
- "* $H_0$:$\\rho_{p}=0$\n",
- "* $H_A$:$\\rho_{p}\\ne0$"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "from statsmodels.stats.diagnostic import acorr_breusch_godfrey\n",
- "\n",
- "acorr_breusch_godfrey(res_h, nlags=1)"
+ "ショックの直接的な影響だけを考えると,比較的にAD曲線の変動幅が大きいことを示している。\n",
+ "次章では,これらの値を使い定量分析を更に進めていく。"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "jp-MarkdownHeadingCollapsed": true
+ },
"source": [
- "上から次の値となる。\n",
- "* $LM$検定統計量\n",
- "* $LM$検定統計量に関する$p$値\n",
- "* $F$検定統計量\n",
- "* $F$検定統計量に関する$p$値"
+ "## 付録"
]
},
{
"cell_type": "markdown",
- "metadata": {},
- "source": [
- "5%の有意水準で帰無仮説を棄却できる。即ち,残差の自己相関があるようだ。\n",
- "\n",
- "次に,Ljung-Box検定もおこなってみよう。\n",
- "\n",
- "<帰無仮説>\n",
- "\n",
- "* $H_0$:残差は独立分布\n",
- "* $H_A$:残差は独立分布ではない"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
+ "metadata": {
+ "jp-MarkdownHeadingCollapsed": true
+ },
"source": [
- "from statsmodels.stats.diagnostic import acorr_ljungbox\n",
- "\n",
- "acorr_ljungbox(res_h.resid, lags=1)"
+ "(sec:18-1-appendix_A)=\n",
+ "### 付録A"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "左の値が検定統計量であり,右の値が$p$値となる。5%の有意水準で帰無仮説を棄却される。自己相関が疑われる結果である。"
+ "この付録は次の2点を示すことを目的としている。\n",
+ "* 自然産出量と物価水準のトレンドは成長し変動するように本文のADASモデルを拡張する。\n",
+ "* 拡張モデルの下でも式[](eq:18-as_small)と式[](eq:18-ad_small)が導出される。"
]
},
{
@@ -1657,34 +2179,57 @@
"jp-MarkdownHeadingCollapsed": true
},
"source": [
- "#### 均一分散の検定"
+ "#### 拡張モデル"
]
},
{
- "cell_type": "code",
- "execution_count": null,
+ "cell_type": "markdown",
"metadata": {},
- "outputs": [],
"source": [
- "from statsmodels.stats.api import het_breuschpagan, het_white"
+ "物価水準のトレンドの式は\n",
+ "\n",
+ "$$\n",
+ "P_t^{*}=P_{t-1}^{*}+aY_t^{*} + A_t+V_t\n",
+ "\\tag{a1}\n",
+ "$$\n",
+ "\n",
+ "で与えられ,次式で自然産出量が決定されると仮定する。\n",
+ "\n",
+ "$$\n",
+ "Y_t^{*}=B_{t}-cP_t^{*} + U_t\n",
+ "\\tag{a2}\n",
+ "$$"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "式[](eq:18-regression-h)を回帰式(定数項あり)として推定した結果`res_h`を使い、均一分散の検定をおこなう。\n",
- "* 帰無仮説:残差は均一分散\n",
- "* 対立仮説:帰無仮説は成立しない"
+ "ここで\n",
+ "* $A_t=$ $P_t^{*}$の増加を決定する項\n",
+ "* $V_t=$ $P_t^{*}$に影響を及ぼすランダムなショック\n",
+ "* $B_t=$ $Y_t^{*}$の増加を決定する項\n",
+ "* $U_t=$ $Y_t^{*}$に影響を及ぼすランダムなショック"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`het_breuschpagan()`と`het_white()`について\n",
- "* 第1引き数:誤差項の値(`res_h.resid`)\n",
- "* 第2引き数:定数項を含む説明変数の値(`res_h.model.exog`)"
+ "一方で,AS曲線とAD曲線を次のように仮定する。\n",
+ "\n",
+ "$$\n",
+ "P_t=P_{t-1}+aY_t + A_t+V_t+v_t\n",
+ "\\tag{a3}\n",
+ "$$\n",
+ "\n",
+ "$$\n",
+ "Y_t=B_{t}-cP_t + U_t+u_t\n",
+ "\\tag{a4}\n",
+ "$$\n",
+ "\n",
+ "* $V_t$は$P_t^{*}$に影響を与えるという意味で,$P_t$に対する恒常的な需要ショックと解釈できる。\n",
+ "* $U_t$は$Y_t^{*}$に影響を与えるという意味で,$Y_t$に対する恒常的な需要ショックと解釈できる。"
]
},
{
@@ -1693,33 +2238,39 @@
"jp-MarkdownHeadingCollapsed": true
},
"source": [
- "##### Breusch-Pagan検定"
+ "#### 式[](eq:18-as_small)と[](eq:18-ad_small)の導出"
]
},
{
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "tags": [
- "remove-cell"
- ]
- },
- "outputs": [],
+ "cell_type": "markdown",
+ "metadata": {},
"source": [
- "het_breuschpagan(res_h.resid, res_h.model.exog)"
+ "式(a1)の両辺から式(a3)を引くと次式となる。\n",
+ "\n",
+ "$$\n",
+ "\\begin{aligned}\n",
+ "P_t-P_t^{*}&=P_{t-1}-P_{t-1}^{*}+a(Y_t-Y_t^{*}) +v_t\\\\\n",
+ "p_t&=p_{t-1}+ay_t+v_t\n",
+ "\\end{aligned}\n",
+ "$$\n",
+ "\n",
+ "2行目は式[](eq:18-as_small)と同じである。"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "返り値(上から)\n",
- "* `LM statistic`:$LM$統計量\n",
- "* `LM p-value`:$LM$の$p$値\n",
- "* `F statistics`:$F$統計量\n",
- "* `F p-value`:$F$の$p$値\n",
+ "次に式(a2)の両辺から式(a4)を引くと次式となる。\n",
"\n",
- "`1`%の有意水準で帰無仮説を棄却できる。"
+ "$$\n",
+ "\\begin{aligned}\n",
+ "Y_t-Y_t^{*}&=-c(P_t-P_t^{*}) + u_t\\\\\n",
+ "y_t&=-cp_t + u_t\n",
+ "\\end{aligned}\n",
+ "$$\n",
+ "\n",
+ "2行目は式[](eq:18-ad_small)と同じとなっている。"
]
},
{
@@ -1728,73 +2279,107 @@
"jp-MarkdownHeadingCollapsed": true
},
"source": [
- "##### White検定"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "tags": [
- "remove-cell"
- ]
- },
- "outputs": [],
- "source": [
- "het_white(res_h.resid, res_h.model.exog)"
+ "#### $Y_t^{*}$と$P_t^{*}$について"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "返り値(上から)\n",
- "* `LM statistic`:$LM$統計量\n",
- "* `LM p-value`:$LM$の$p$値\n",
- "* `F statistics`:$F$統計量\n",
- "* `F p-value`:$F$の$p$値\n",
+ "次の仮定を置こう。\n",
"\n",
- "`1`%の有意水準で帰無仮説を棄却できる。"
+ "$$\n",
+ "A_t=-aY_{t-1}^{*}\n",
+ "$$\n",
+ "\n",
+ "1期遅れで自然産出量は$P_t^{*}$を減少させる。\n",
+ "自然産出量の増加は生産性の上昇に起因するものであり,それにより物価水準とそのトレンドは減少するというメカニズムを捉えている。\n",
+ "\n",
+ "この仮定を使い,式(a3)を次の様に書き換えよう。\n",
+ "\n",
+ "$$\n",
+ "P_t^{*}-P_{t-1}^{*}=aY_t^{*} - aY_{t-1}^{*}+V_t\n",
+ "$$\n",
+ "\n",
+ "$$\n",
+ "\\Delta P_t^{*}=a\\Delta Y_t^{*} + V_t\n",
+ "\\tag{a5}\n",
+ "$$\n",
+ "\n",
+ "* $\\Delta P_t^{*}=P_t^{*}-P_{t-1}^{*}$: 物価水準のトレンドの増加率\n",
+ "* $\\Delta Y_t^{*}=Y_t^{*}-Y_{t-1}^{*}$: 自然産出量の増加率\n",
+ "\n",
+ "この式は,物価水準トレンドの増加率$\\Delta P_t^{*}$は自然産出量の成長率と比例し,$V_t$によりランダムに変動することを示しているの。"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "不均一分散の可能性が高いと判断できる。"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "jp-MarkdownHeadingCollapsed": true
- },
- "source": [
- "#### `h`の不均一分散自己相関頑健推定"
+ "次に式(a4)の差分を次の様に計算しよう。\n",
+ "\n",
+ "$$\n",
+ "Y_t^{*}-Y_{t-1}^{*}=B_t-B_{t-1}-c(P_t^{*}-P_{t-1}^{*}) + U_t-U_{t-1}\n",
+ "$$\n",
+ "\n",
+ "$$\n",
+ "\\Delta Y_t=\\Delta B_t-c\\Delta P_t^{*} + \\Delta U_t\n",
+ "\\tag{a6}\n",
+ "$$\n",
+ "\n",
+ "* $\\Delta B_t=B_t-B_{t-1}$: 生産性の増加率と解釈する(例えば,`0.02`の様な定数にすると毎期`2`%の成長率となる)\n",
+ "* $\\Delta U_t=U_t-U_{t-1}$: 恒常的な需要ショックの差分"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`res_h`を使い、`h`の不均一分散自己相関頑健推定ををおこなう。詳しい説明は[Pythonで学ぶ入門計量経済学ー不均一分散](https://py4etrics.github.io/14_Hetero.html)を参照しよう。"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "res_h_robust = res_h.get_robustcov_results(cov_type='HAC', maxlags=1, use_t=True)\n",
- "print(res_h_robust.summary().tables[1])"
+ "式(a5)を式(a6)に代入し整理すると次式となる。\n",
+ "\n",
+ "$$\n",
+ "\\begin{aligned}\n",
+ "\\Delta Y_t&=\\Delta B_t-c[a\\Delta Y_t^{*}+V_t] + \\Delta U_t\\\\\n",
+ "(1+ac)\\Delta Y_t&=\\Delta B_t-cV_t + \\Delta U_t\\\\\n",
+ "&\\Downarrow\n",
+ "\\end{aligned}\n",
+ "$$\n",
+ "\n",
+ "$$\n",
+ "\\Delta Y_t=\\frac{\\Delta B_t-cV_t + \\Delta U_t}{1+ac}\n",
+ "\\tag{a7}\n",
+ "$$"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`deflator_cycle_lag`の推定値の統計的優位性は高いことが確認できる。"
+ "式(a7)は自然産出量の成長率を表しており,ランダムなショックにより,変動しながら増加することになる。ショックがない定常状態($V_t=U_t=U_{t=1}0$)では\n",
+ "\n",
+ "$$\n",
+ "\\Delta Y_t=\\frac{\\Delta B_t}{1+ac}\n",
+ "\\tag{a8}\n",
+ "$$\n",
+ "\n",
+ "の率で増加することになる。また式(a8)を式(a5)に代入すると次式となる。\n",
+ "\n",
+ "$$\n",
+ "\\begin{align}\n",
+ "\\Delta P_t^{*}\n",
+ "&=a\\frac{\\Delta B_t-cV_t + \\Delta U_t}{1+ac} + V_t\\\\\n",
+ "&=a\\frac{\\Delta B_t+V_t + \\Delta U_t}{1+ac}\n",
+ "\\end{align}\n",
+ "$$\n",
+ "\n",
+ "自然産出量と同様に,変動しつつ増加することになる。ショックがない定常状態では\n",
+ "\n",
+ "$$\n",
+ "\\Delta P_t^{*}\n",
+ "=\\frac{a}{1+ac}\\Delta B_t\n",
+ "$$\n",
+ "\n",
+ "の率で増加することになる。"
]
},
{
@@ -1807,52 +2392,27 @@
"### 付録B"
]
},
- {
- "cell_type": "markdown",
- "metadata": {
- "jp-MarkdownHeadingCollapsed": true
- },
- "source": [
- "#### 均一分散の検定"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "推定結果`res_d`を使い付録Aと同じ方法で検定をおこなう。"
- ]
- },
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "##### Breusch-Pagan検定"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": [
- "het_breuschpagan(res_d.resid, res_d.model.exog)"
+ "#### 残差の自己相関の検定"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "有意水準`5`%で帰無仮説を棄却できるが、`1`%では棄却できない。"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "jp-MarkdownHeadingCollapsed": true
- },
- "source": [
- "##### White検定"
+ "Breusch-Godfrey検定を考えよう。\n",
+ "\n",
+ "<帰無仮説>\n",
+ "\n",
+ "$$\n",
+ "e_{pt}=\\rho_{p}e_{pt-1}+e_{pt}\n",
+ "$$\n",
+ "\n",
+ "* $H_0$:$\\rho_{p}=0$\n",
+ "* $H_A$:$\\rho_{p}\\ne0$"
]
},
{
@@ -1861,35 +2421,34 @@
"metadata": {},
"outputs": [],
"source": [
- "het_white(res_d.resid, res_d.model.exog)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "有意水準`10`%で帰無仮説を棄却できるが、`5`%では棄却できない。"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "均一分散に十分に自信が持てる結果ではない。"
+ "from statsmodels.stats.diagnostic import acorr_breusch_godfrey\n",
+ "\n",
+ "acorr_breusch_godfrey(res_h, nlags=1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "#### `d`の不均一分散自己相関頑健推定"
+ "上から次の値となる。\n",
+ "* $LM$検定統計量\n",
+ "* $LM$検定統計量に関する$p$値\n",
+ "* $F$検定統計量\n",
+ "* $F$検定統計量に関する$p$値"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`res_d`を使い、`d`の不均一分散自己相関頑健推定ををおこなう。詳しい説明は[Pythonで学ぶ入門計量経済学ー不均一分散](https://py4etrics.github.io/14_Hetero.html)を参照しよう。"
+ "5%の有意水準で帰無仮説を棄却できる。即ち,残差の自己相関があるようだ。\n",
+ "\n",
+ "次に,Ljung-Box検定もおこなってみよう。\n",
+ "\n",
+ "<帰無仮説>\n",
+ "\n",
+ "* $H_0$:残差は独立分布\n",
+ "* $H_A$:残差は独立分布ではない"
]
},
{
@@ -1898,15 +2457,16 @@
"metadata": {},
"outputs": [],
"source": [
- "res_d_hac = smf.ols('gdp_cycle ~ deflator_cycle_lag', data=df).fit(cov_type='HAC', cov_kwds={'maxlags':1}, use_t=True)\n",
- "print(res_d_hac.summary().tables[1])"
+ "from statsmodels.stats.diagnostic import acorr_ljungbox\n",
+ "\n",
+ "acorr_ljungbox(res_h.resid, lags=1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
- "`deflator_cycle_lag`の推定値の統計的優位性は高いことが確認できる。"
+ "左の値が検定統計量であり,右の値が$p$値となる。5%の有意水準で帰無仮説を棄却される。自己相関が疑われる結果である。"
]
},
{
@@ -1946,7 +2506,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.12.2"
+ "version": "3.12.7"
},
"nteract": {
"version": "0.15.0"
diff --git a/18_ADAS-2.ipynb b/18_ADAS-2.ipynb
index ee705cc1..dbdd7f53 100644
--- a/18_ADAS-2.ipynb
+++ b/18_ADAS-2.ipynb
@@ -21,7 +21,7 @@
},
{
"cell_type": "code",
- "execution_count": 47,
+ "execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
@@ -80,7 +80,7 @@
},
{
"cell_type": "code",
- "execution_count": 48,
+ "execution_count": 2,
"metadata": {
"hidden": true,
"tags": [
@@ -89,15 +89,15 @@
},
"outputs": [],
"source": [
- "ahat = 0.4405822671569939\n",
- "chat = 0.5002776200207564\n",
- "u_std = 0.01475683470914417\n",
- "v_std = 0.007494205944060558"
+ "ahat = 0.82274785974591313487\n",
+ "chat = 0.26898029723198069174\n",
+ "v_std = 0.01254675097186065853\n",
+ "u_std = 0.01463393441548518220"
]
},
{
"cell_type": "code",
- "execution_count": 50,
+ "execution_count": 3,
"metadata": {
"hidden": true
},
@@ -106,18 +106,18 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "aの値:0.441\n",
- "cの値:0.500\n",
- "uの標準偏差:0.014757\n",
- "vの標準偏差:0.007494\n"
+ "aの値:0.823\n",
+ "cの値:0.269\n",
+ "vの標準偏差:0.012547\n",
+ "uの標準偏差:0.014634\n"
]
}
],
"source": [
"print(f'aの値:{ahat:.3f}')\n",
"print(f'cの値:{chat:.3f}')\n",
- "print(f'uの標準偏差:{u_std:.6f}')\n",
- "print(f'vの標準偏差:{v_std:.6f}')"
+ "print(f'vの標準偏差:{v_std:.6f}')\n",
+ "print(f'uの標準偏差:{u_std:.6f}')"
]
},
{
@@ -325,7 +325,7 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": 6,
"metadata": {
"hidden": true
},
@@ -377,14 +377,14 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": 7,
"metadata": {
"hidden": true
},
"outputs": [
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
"