Make frequency filtering robust to missing data

src/noisepy/seis/S1_fft_cc_MPI.py

@@ -291,8 +291,11 @@ def _filter_channel_data(
     tuples: List[Tuple[Channel, ChannelData]], samp_freq: int
 ) -> List[Tuple[Channel, ChannelData]]:
     frequencies = set(t[1].sampling_rate for t in tuples)
+    frequencies = list(filter(lambda f: f >= samp_freq, frequencies))
+    if len(frequencies) == 0:
+        return []
     closest_freq = min(
-        filter(lambda f: f >= samp_freq, frequencies),
+        frequencies,
         key=lambda f: max(f - samp_freq, 0),
     )
     filtered_tuples = list(filter(lambda tup: tup[1].sampling_rate == closest_freq, tuples))