Update test_convergence and lambda tests in convergence.moving_average

src/alchemlyb/convergence/convergence.py

@@ -98,14 +98,13 @@ def forward_backward_convergence(
 
     # Check that each df in the list has only one value of lambda
     for i, df in enumerate(df_list):
-        lambda_values = np.unique([x[1] for x in df.index.to_numpy()])
+        lambda_values = list(set([x[1:] for x in df.index.to_numpy()]))
         if len(lambda_values) > 1:
-            if len(df_list[0].index[0]) > 2:
-                lambda_values = np.unique([x[2] for x in df.index.to_numpy()])
-                if len(lambda_values) > 1:
-                    raise ValueError("Provided DataFrame, df_list[{}] has more than one lambda value".format(i))
-            else:
-                raise ValueError("Provided DataFrame, df_list[{}] has more than one lambda value".format(i))
+            for j in range(len(lambda_values[0])):
+                if len(list(set([x[j] for x in lambda_values]))) > 1:
+                    raise ValueError(
+                        "Provided DataFrame, df_list[{}] has more than one lambda value in df.index[{}]".format(i, j)
+                    )
 
     logger.info("Begin forward analysis")
     forward_list = []
@@ -459,18 +458,17 @@ def moving_average(df_list, estimator="MBAR", num=10, **kwargs):
 
     # Check that each df in the list has only one value of lambda
     for i, df in enumerate(df_list):
-        lambda_values = np.unique([x[1] for x in df.index.to_numpy()])
+        lambda_values = list(set([x[1:] for x in df.index.to_numpy()]))
         if len(lambda_values) > 1:
-            if len(df_list[0].index[0]) > 2:
-                lambda_values = np.unique([x[2] for x in df.index.to_numpy()])
-                if len(lambda_values) > 1:
-                    raise ValueError("Provided DataFrame, df_list[{}] has more than one lambda value in df.index[2]".format(i))
-            else:
-                raise ValueError("Provided DataFrame, df_list[{}] has more than one lambda value in df.index[1]".format(i))
+            for j in range(len(lambda_values[0])):
+                if len(list(set([x[j] for x in lambda_values]))) > 1:
+                    raise ValueError(
+                        "Provided DataFrame, df_list[{}] has more than one lambda value in df.index[{}]".format(i, j)
+                    )
 
     if estimator in ["BAR"] and len(df_list) > 2:
         raise ValueError(
-            "Restrict to two DataFrames, one with a fep-lambda value and one its forward adjacent state for a"
+            "Restrict to two DataFrames, one with a fep-lambda value and one its forward adjacent state for a "
             "meaningful result."
         )
 

src/alchemlyb/tests/test_convergence.py

@@ -2,6 +2,7 @@
 import pandas as pd
 import pytest
 
+from alchemlyb import concat
 from alchemlyb.convergence import (
     forward_backward_convergence,
     fwdrev_cumavg_Rc,
@@ -52,31 +53,31 @@ def test_moving_average_error_2_mbar(gmx_ABFE_complex_u_nk, estimator):
     df_list = gmx_ABFE_complex_u_nk[10:15]
     with pytest.raises(
         ValueError,
-        match=r"Restrict to a single fep-lambda value for a meaningful result. .*",
+        match=r"Provided DataFrame, df_list\[0\] has more than one lambda value in df.index\[0\]",
     ):
-        _ = moving_average(df_list, estimator)
+        _ = moving_average([concat(df_list)], estimator)
 
     df_list = gmx_ABFE_complex_u_nk[14:17]
     with pytest.raises(
         ValueError,
-        match=r"Restrict to a single fep-lambda value for a meaningful result. .*",
+        match=r"Provided DataFrame, df_list\[0\] has more than one lambda value in df.index\[1\]",
     ):
-        _ = moving_average(df_list, estimator)
+        _ = moving_average([concat(df_list)], estimator)
 
 
 @pytest.mark.parametrize("estimator", ["BAR"])
 def test_moving_average_error_2_bar(gmx_ABFE_complex_u_nk, estimator):
     df_list = gmx_ABFE_complex_u_nk[10:13]
     with pytest.raises(
         ValueError,
-        match=r"Restrict to a fep-lambda value and its forward adjacent state .*",
+        match=r"Restrict to two DataFrames, one with a fep-lambda value .*",
     ):
         _ = moving_average(df_list, estimator)
 
     df_list = gmx_ABFE_complex_u_nk[14:17]
     with pytest.raises(
         ValueError,
-        match=r"Restrict to a fep-lambda value and its forward adjacent state .*",
+        match=r"Restrict to two DataFrames, one with a fep-lambda value .*",
     ):
         _ = moving_average(df_list, estimator)