unittests and debug for sinusoidal stimulus

stimuli/SinusAmplitudeModulation.py

@@ -47,6 +47,40 @@ class SinusAmplitudeModulationStimulus(AbstractStimulus):
 
 # @jit(nopython=True)  # makes it slower?
 def convert_to_array(carrier_freq, amplitude, modulation_freq, contrast, start_time, duration, time_start, total_time, step_size_s):
+    full_time = np.arange(time_start, time_start + total_time, step_size_s)
+    full_carrier = np.sin(2 * np.pi * carrier_freq * full_time)
+    if start_time > time_start+duration or start_time+duration < time_start:
+        return full_carrier * amplitude
+    else:
+        if start_time >= time_start:
+            am_start = start_time
+        else:
+            am_start = time_start
+
+        if time_start + total_time >= start_time + duration:
+            am_end = start_time + duration
+        else:
+            am_end = time_start + total_time
+
+
+        idx_start = (am_start - time_start) / step_size_s
+        idx_end = (am_end - time_start) / step_size_s
+
+        if idx_start != round(idx_start) or idx_end != round(idx_end):
+            raise ValueError("Didn't calculate integers when searching the start and end index. start:", idx_start, "end:", idx_end)
+        # print("am_start: {:.0f}, am_end: {:.0f}, length: {:.0f}".format(am_start, am_end, am_end-am_start))
+
+        idx_start = int(idx_start)
+        idx_end = int(idx_end)
+
+        am = 1 + contrast * np.sin(2 * np.pi * modulation_freq * full_time[idx_start:idx_end])
+
+        values = full_carrier * amplitude
+        values[idx_start:idx_end] = values[idx_start:idx_end]*am
+
+        return values
+
+
     # if the whole stimulus time has the amplitude modulation just built it at once;
     if time_start >= start_time and start_time+duration < time_start+total_time:
         carrier = np.sin(2 * np.pi * carrier_freq * np.arange(start_time, total_time - start_time, step_size_s))
@@ -55,8 +89,9 @@ def convert_to_array(carrier_freq, amplitude, modulation_freq, contrast, start_t
         return values
 
     # if it is split into parts with and without amplitude modulation built it in parts:
-    values = np.array([], dtype='float32')
+    values = np.array([])
 
+    # there is some time before the modulation starts:
     if time_start < start_time:
         carrier_before_am = np.sin(2 * np.pi * carrier_freq * np.arange(time_start, start_time, step_size_s))
         values = np.concatenate((values, amplitude * carrier_before_am))