Merge pull request #38 from Lucs1590/22-add-multi-language-descriptio…

…n-support

Add multi language description support

requirements.txt

@@ -6,6 +6,6 @@ numpy==1.26.4
 pandas==2.2.3
 python-dotenv==1.0.1
 questionary==2.0.1
-scipy==1.14.1
+scipy==1.13.1
 tcxreader==0.4.10
 tqdm==4.67.0

setup.py

@@ -14,7 +14,7 @@
             "pandas==2.2.3",
             "python-dotenv==1.0.1",
             "questionary==2.0.1",
-            "scipy==1.14.1",
+            "scipy==1.13.1",
             "tcxreader==0.4.10",
             "tqdm==4.67.0"
     ],

src/main.py

@@ -61,8 +61,9 @@ def main():
             _, tcx_data = validate_tcx_file(file_path)
             if ask_llm_analysis():
                 plan = ask_training_plan()
+                language = ask_desired_language()
                 logger.info("Performing LLM analysis")
-                perform_llm_analysis(tcx_data, sport, plan)
+                perform_llm_analysis(tcx_data, sport, plan, language)
         else:
             logger.error("Invalid sport selected")
             raise ValueError("Invalid sport selected")
@@ -194,35 +195,46 @@ def ask_training_plan() -> str:
     ).ask()
 
 
-def perform_llm_analysis(data: TCXReader, sport: str, plan: str) -> str:
-    dataframe = preprocess_trackpoints_data(data)
+def ask_desired_language() -> str:
+    return questionary.text(
+        "In which language do you want the analysis to be provided? (Default is Portuguese)",
+        default="Portuguese (Brazil)"
+    ).ask()
 
-    prompt_template = """
-    SYSTEM: You are an AI coach helping athletes optimize and improve their performance. 
-    Based on the provided {sport} training session data, perform the following analysis:
 
-    1. Identify key performance metrics.
-    2. Highlight the athlete's strengths during the session.
-    3. Pinpoint areas where the athlete can improve.
-    4. Offer actionable suggestions for enhancing performance in future {sport} sessions.
+def perform_llm_analysis(data: TCXReader, sport: str, plan: str, language: str) -> str:
+    dataframe = preprocess_trackpoints_data(data)
 
-    Training session data:
+    prompt_template = """
+    SYSTEM: You are an AI performance coach specializing in analyzing athletic performance to help athletes with their trainings.
+    Using the provided {sport} training session data, analyze the athlete's performance and deliver a detailed analysis and practical advice in {language} language.
+    Your analysis should include:
+
+    1. Key Performance Metrics: Identify and Evaluate the most relevant metrics from the session, understanding the athlete's overall performance
+    2. Strengths: Highlight the athlete's strongest aspects during the session, supported by specific metrics.
+    3. Improvement Opportunities: Pinpoint specific areas for growth and improvement.
+    4. Actionable Suggestions: Provide clear, practical recommendations to help the athlete enhance their performance in future {sport} sessions.
+    
+    Ensure your response is data-driven, clear, and motivational, helping the athlete make measurable progress.
+    
+    Training Session Data:
     {training_data}
     """
 
     if plan:
-        prompt_template += "\nTraining plan details: {plan}"
+        prompt_template += "\n\nTraining Plan Details:\n{plan}"
 
     prompt = PromptTemplate.from_template(prompt_template).format(
         sport=sport,
         training_data=dataframe.to_csv(index=False),
+        language=language,
         plan=plan
     )
 
     openai_llm = ChatOpenAI(
         openai_api_key=os.getenv("OPENAI_API_KEY"),
-        model_name="gpt-4o",
-        max_tokens=1500,
+        model_name="gpt-4o-mini",
+        max_tokens=2000,
         temperature=0.6,
         max_retries=5
     )

tests/test_main.py

@@ -24,6 +24,7 @@
     get_latest_download,
     validation,
     ask_training_plan,
+    ask_desired_language,
     ask_llm_analysis,
     perform_llm_analysis,
     preprocess_trackpoints_data,
@@ -198,6 +199,7 @@ def test_main_invalid_sport(self, mock_indent, mock_validate, mock_format, mock_
         mock_validate.assert_not_called()
         mock_indent.assert_not_called()
 
+    @patch('src.main.ask_desired_language')
     @patch('src.main.ask_training_plan')
     @patch('src.main.perform_llm_analysis')
     @patch('src.main.ask_llm_analysis')
@@ -211,14 +213,15 @@ def test_main_invalid_sport(self, mock_indent, mock_validate, mock_format, mock_
     @patch('src.main.indent_xml_file')
     def test_main_bike_sport(self, mock_indent, mock_validate, mock_format, mock_ask_path, mock_download,
                              mock_ask_id, mock_ask_location, mock_ask_sport, mock_llm_analysis, mock_perform_llm,
-                             mock_training_plan):
+                             mock_training_plan, mock_language):
         mock_ask_sport.return_value = "Bike"
         mock_ask_location.return_value = "Local"
         mock_ask_path.return_value = "assets/bike.tcx"
         mock_llm_analysis.return_value = True
         mock_validate.return_value = True, "TCX Data"
         mock_perform_llm.return_value = "Training Plan"
         mock_training_plan.return_value = ""
+        mock_language.return_value = "Portuguese"
 
         main()
 
@@ -306,6 +309,16 @@ def test_ask_training_plan(self):
             )
             self.assertEqual(result, "")
 
+    def test_ask_desired_language(self):
+        with patch('src.main.questionary.text') as mock_text:
+            mock_text.return_value.ask.return_value = "Portuguese"
+            result = ask_desired_language()
+            mock_text.assert_called_once_with(
+                'In which language do you want the analysis to be provided? (Default is Portuguese)',
+                default='Portuguese (Brazil)'
+            )
+            self.assertEqual(result, "Portuguese")
+
     def test_ask_llm_analysis(self):
         with patch('src.main.questionary.confirm') as mock_confirm:
             mock_confirm.return_value.ask.return_value = True
@@ -323,8 +336,9 @@ def test_perform_llm_analysis(self, mock_chat):
         tcx_data = self.running_example_data
         sport = "Run"
         plan = "Training Plan"
+        lang = "Portuguese"
 
-        result = perform_llm_analysis(tcx_data, sport, plan)
+        result = perform_llm_analysis(tcx_data, sport, plan, lang)
         self.assertEqual(result, "Training Plan")
 
     def test_preprocess_running_trackpoints_data(self):