[Fix] build docs warning

neurokit2/__init__.py

@@ -1,4 +1,5 @@
 """Top-level package for NeuroKit."""
+
 import datetime
 import platform
 

neurokit2/complexity/utils_complexity_symbolize.py

@@ -184,8 +184,8 @@ def complexity_symbolize(signal, method="mean", c=3, random_state=None, show=Fal
             symbolic = (signal > np.nanmean(signal)).astype(int)
             if show is True:
                 df = pd.DataFrame({"A": signal, "B": signal})
-                df["A"][df["A"] > np.nanmean(signal)] = np.nan
-                df["B"][df["B"] <= np.nanmean(signal)] = np.nan
+                df.loc[df["A"] > np.nanmean(signal), "A"] = np.nan
+                df.loc[df["B"] <= np.nanmean(signal), "B"] = np.nan
                 df.plot()
                 plt.axhline(y=np.nanmean(signal), color="r", linestyle="dotted")
                 plt.title("Method A")
@@ -194,8 +194,8 @@ def complexity_symbolize(signal, method="mean", c=3, random_state=None, show=Fal
             symbolic = (signal > np.nanmedian(signal)).astype(int)
             if show is True:
                 df = pd.DataFrame({"A": signal, "B": signal})
-                df["A"][df["A"] > np.nanmedian(signal)] = np.nan
-                df["B"][df["B"] <= np.nanmedian(signal)] = np.nan
+                df.loc[df["A"] > np.nanmedian(signal), "A"] = np.nan
+                df.loc[df["B"] <= np.nanmedian(signal), "B"] = np.nan
                 df.plot()
                 plt.axhline(y=np.nanmean(signal), color="r", linestyle="dotted")
                 plt.title("Binarization by median")
@@ -206,8 +206,9 @@ def complexity_symbolize(signal, method="mean", c=3, random_state=None, show=Fal
             symbolic = np.logical_or(signal < m - sd, signal > m + sd).astype(int)
             if show is True:
                 df = pd.DataFrame({"A": signal, "B": signal})
-                df["A"][np.logical_or(signal < m - sd, signal > m + sd)] = np.nan
-                df["B"][~np.isnan(df["A"])] = np.nan
+                condition = np.logical_or(signal < m - sd, signal > m + sd)
+                df.loc[condition, "A"] = np.nan
+                df.loc[~np.isnan(df["A"]), "B"] = np.nan
                 df.plot()
                 plt.axhline(y=m - sd, color="r", linestyle="dotted")
                 plt.axhline(y=m + sd, color="r", linestyle="dotted")
@@ -217,8 +218,8 @@ def complexity_symbolize(signal, method="mean", c=3, random_state=None, show=Fal
             symbolic = np.signbit(np.diff(signal)).astype(int)
             if show is True:
                 df = pd.DataFrame({"A": signal, "B": signal})
-                df["A"][np.insert(symbolic, 0, False)] = np.nan
-                df["B"][~np.isnan(df["A"])] = np.nan
+                df.loc[np.insert(symbolic, 0, False), "A"] = np.nan
+                df.loc[~np.isnan(df["A"]), "B"] = np.nan
                 df.plot()
                 plt.title("Method C")
 

neurokit2/emg/emg_activation.py

@@ -231,16 +231,16 @@ def emg_activation(
 
     # Modify output produced by signal_formatpeaks.
     for x in range(len(emg_amplitude)):
-        if df_activity["EMG_Activity"][x] != 0:
+        if df_activity.loc[x, "EMG_Activity"] != 0:
             if df_activity.index[x] == df_activity.index.get_loc(x):
-                df_activity["EMG_Activity"][x] = 1
+                df_activity.loc[x, "EMG_Activity"] = 1
             else:
-                df_activity["EMG_Activity"][x] = 0
-        if df_offsets["EMG_Offsets"][x] != 0:
+                df_activity.loc[x, "EMG_Activity"] = 0
+        if df_offsets.loc[x, "EMG_Offsets"] != 0:
             if df_offsets.index[x] == df_offsets.index.get_loc(x):
-                df_offsets["EMG_Offsets"][x] = 1
+                df_offsets.loc[x, "EMG_Offsets"] = 1
             else:
-                df_offsets["EMG_Offsets"][x] = 0
+                df_offsets.loc[x, "EMG_Offsets"] = 0
 
     activity_signal = pd.concat([df_activity, df_onsets, df_offsets], axis=1)
 

neurokit2/signal/signal_resample.py

@@ -157,11 +157,11 @@ def _resample_poly(signal, desired_length):
 
 def _resample_pandas(signal, desired_length):
     # Convert to Time Series
-    index = pd.date_range("20131212", freq="L", periods=len(signal))
+    index = pd.date_range("20131212", freq="ms", periods=len(signal))
     resampled_signal = pd.Series(signal, index=index)
 
     # Create resampling factor
-    resampling_factor = str(np.round(1 / (desired_length / len(signal)), 6)) + "L"
+    resampling_factor = str(np.round(1 / (desired_length / len(signal)), 6)) + "ms"
 
     # Resample
     resampled_signal = resampled_signal.resample(resampling_factor).bfill().values