adding a fallback on preliminary calculation to avoid IndexError at s…

…hort source receiver distances, allowing for window selection rather than code crash, also added additional log messages for this preliminaries calculation. See related #21

pyflex/window_selector.py

@@ -271,7 +271,15 @@ def calculate_preliminiaries(self):
         if not len(self.peaks) and len(self.troughs):
             return
 
+        # Try to use theoretical arrival times to curtail the bounds of the
+        # troughs, which should slim down the total number of window choices
+        # that can be considered, this doesn't always work so it has a fallback
+        # 
         if self.ttimes:
+            # Retain original peaks and troughs incase this doesn't work
+            _peaks = self.peaks
+            _troughs = self.troughs
+
             offset = self.event.origin_time - self.observed.stats.starttime
             min_time = self.ttimes[0]["time"] - \
                 self.config.max_time_before_first_arrival + offset
@@ -290,8 +298,11 @@ def calculate_preliminiaries(self):
             self.troughs = self.troughs[(self.troughs >= min_idx) &
                                         (self.troughs <= max_idx)]
 
-            # If troughs have been removed, readd them add the boundaries.
-            if len(self.troughs):
+            # If the first or last trough was removed from the rejected points,
+            # replace them so we know our boundaries
+            if self.troughs:
+                logger.debug("Using theoretical first arrival %.2fs to curtail "
+                             "STA/LTA selection" % min_time)
                 if first_trough != self.troughs[0]:
                     self.troughs = np.concatenate([
                         np.array([min_idx], dtype=self.troughs.dtype),
@@ -300,10 +311,18 @@ def calculate_preliminiaries(self):
                     self.troughs = np.concatenate([
                         self.troughs,
                         np.array([max_idx], dtype=self.troughs.dtype)])
-            # Make sure peaks are inside the troughs!
-            min_trough, max_trough = self.troughs[0], self.troughs[-1]
-            self.peaks = self.peaks[(self.peaks > min_trough) &
-                                    (self.peaks < max_trough)]
+
+                # Make sure peaks are inside the troughs!
+                min_trough, max_trough = self.troughs[0], self.troughs[-1]
+                self.peaks = self.peaks[(self.peaks > min_trough) &
+                                        (self.peaks < max_trough)]
+            # Edge case where this processing removes all potential troughs. 
+            # This often occurs when source-receiver distance is short w.r.t 
+            # wavelengths. In this case we just move on like this never 
+            # happend
+            else:
+                self.peaks = _peaks
+                self.troughs = _troughs
 
     def select_windows(self):
         """