added test that filter complex response does not choke when provided …

…with a simple integer or float rather than a numpy array -- issue #18

mt_metadata/timeseries/filters/test_filters.py

@@ -83,6 +83,7 @@ def test_filter_generation_from_xml_via_obspy(inventory):
                 for fltr in filters_list:
                     print(fltr.type)
                     frequencies = np.logspace(-3, 3, 200)
+                    test_filter_response_works_for_float(fltr)
                     # fltr.plot_complex_response(frequencies)
                     # fltr.plot_response(None, x_units='frequency')
                 channel_response_filter = ChannelResponseFilter(filters_list=filters_list)
@@ -94,6 +95,12 @@ def test_filter_generation_from_xml_via_obspy(inventory):
     print("ok")
 
 
+def test_filter_response_works_for_float(fltr):
+    integer_response = fltr.complex_response(1)
+    float_response = fltr.complex_response(1.0)
+    print(f'integer response {integer_response},  float response {float_response}')
+    return
+
 def test_correct_sense_of_normalization_factor():
     """
     Load in an mth5 with data from NIMS system and calibrate the data.  Confirm whether the normalization factor