remove reliance on NaNs for interpolation

Author: SBlokhuizen

waveform_editor/waveform.py

@@ -86,74 +86,42 @@ def _evaluate_tendencies(self, time, eval_derivatives=False):
         Returns:
             numpy array containing the computed values.
         """
-        values = np.full_like(time, np.nan, dtype=float)
+        values = np.zeros_like(time, dtype=float)
 
-        for tendency in self.tendencies:
+        for i, tendency in enumerate(self.tendencies):
             mask = (time >= tendency.start) & (time <= tendency.end)
             if np.any(mask):
                 if eval_derivatives:
                     values[mask] = tendency.get_derivative(time[mask])
                 else:
                     _, values[mask] = tendency.get_value(time[mask])
 
-        # If there still remain nans in the values, this means that there are gaps
-        # between the tendencies. In this case we linearly interpolate between the gap
-        # values
-        for idx, t in enumerate(time):
-            if np.isnan(values[idx]):
-                # The derivatives of interpolated gaps are not calculated
-                if eval_derivatives:
-                    values[idx] = 0
-                else:
-                    values[idx] = self._interpolate_gap(t)
+            # If there still remain nans in the values, this means that there are gaps
+            # between the tendencies. In this case we linearly interpolate between the
+            # gap values
+            if i and tendency.prev_tendency.end < tendency.start:
+                mask = (time < tendency.start) & (time > tendency.prev_tendency.end)
+                if np.any(mask):
+                    if eval_derivatives:
+                        values[mask] = 0
+                    else:
+                        values[mask] = np.interp(
+                            time[mask],
+                            [tendency.prev_tendency.end, tendency.start],
+                            [tendency.prev_tendency.end_value, tendency.start_value],
+                        )
+        # Handle extrapolation
+        if eval_derivatives:
+            values[time < self.tendencies[0].start] = 0
+            values[time > self.tendencies[-1].end] = 0
+        else:
+            first_tendency = self.tendencies[0]
+            values[time < first_tendency.start] = first_tendency.start_value
 
+            last_tendency = self.tendencies[-1]
+            values[time > last_tendency.end] = last_tendency.end_value
         return values
 
-    def _interpolate_gap(self, t):
-        """Interpolates the value for a given time t based on the nearest tendencies.
-        Also extrapolates the values if the time requested falls before the first, or
-        after the last tendency.
-
-        Args:
-            t: The time for which the value needs to be interpolated.
-
-        Returns:
-            The interpolated value.
-        """
-        # Find nearest tendencies before and after time t
-        prev_tendency = max(
-            (tend for tend in self.tendencies if tend.end <= t),
-            default=None,
-            key=lambda tend: tend.end,
-        )
-        next_tendency = min(
-            (tend for tend in self.tendencies if tend.start >= t),
-            default=None,
-            key=lambda tend: tend.start,
-        )
-
-        if prev_tendency and next_tendency:
-            val_end = prev_tendency.end_value
-            val_start = next_tendency.start_value
-
-            return np.interp(
-                t, [prev_tendency.end, next_tendency.start], [val_end, val_start]
-            )
-
-        # Handle extrapolation if t is before the first or after the last tendency
-        if prev_tendency is None:
-            next_tendency = self.tendencies[0]
-            val_start = next_tendency.start_value
-            return val_start
-
-        if next_tendency is None:
-            prev_tendency = self.tendencies[-1]
-            val_end = prev_tendency.end_value
-            return val_end
-
-        # If no valid interpolation or extrapolation can be performed, return 0
-        return 0.0
-
     def calc_length(self):
         """Returns the length of the waveform."""
         return self.tendencies[-1].end - self.tendencies[0].start