fix handling of nan value in averaging

qiskit_experiments/curve_analysis/curve_analysis.py

@@ -272,7 +272,7 @@ def _format_data(
         average = averaging_methods[self.options.average_method]
         model_names = self.model_names()
         formatted = []
-        for (class_id, xv), g in groupby(sorted(curve_data.values, key=sort_by), key=sort_by):
+        for (_, xv), g in groupby(sorted(curve_data.values, key=sort_by), key=sort_by):
             g_values = np.array(list(g))
             g_dict = dict(zip(columns, g_values.T))
             avg_yval, avg_yerr, shots = average(g_dict["yval"], g_dict["yerr"], g_dict["shots"])

qiskit_experiments/curve_analysis/utils.py

@@ -18,6 +18,7 @@
 import asteval
 import lmfit
 import numpy as np
+import pandas as pd
 from qiskit.utils.deprecation import deprecate_func
 from qiskit.utils import detach_prefix
 from uncertainties import UFloat, wrap as wrap_function
@@ -243,9 +244,9 @@ def shot_weighted_average(
     if len(yvals) == 1:
         return yvals[0], yerrs[0], shots[0]
 
-    if np.any(shots < -1):
+    if any(s is pd.NA for s in shots):
         # Shot number is unknown
-        return np.mean(yvals), np.nan, -1
+        return np.mean(yvals), np.nan, pd.NA
 
     total_shots = np.sum(shots)
     weights = shots / total_shots
@@ -276,7 +277,7 @@ def inverse_weighted_variance(
     if len(yvals) == 1:
         return yvals[0], yerrs[0], shots[0]
 
-    total_shots = np.sum(shots) if all(shots > 0) else -1
+    total_shots = np.sum(shots)
     weights = 1 / yerrs**2
     yvar = 1 / np.sum(weights)
 
@@ -307,7 +308,7 @@ def sample_average(
     if len(yvals) == 1:
         return yvals[0], 0.0, shots[0]
 
-    total_shots = np.sum(shots) if all(shots > 0) else -1
+    total_shots = np.sum(shots)
 
     avg_yval = np.mean(yvals)
     avg_yerr = np.sqrt(np.mean((avg_yval - yvals) ** 2) / len(yvals))