18.09

eigenmannia_code/step_response_eigen.py

@@ -0,0 +1,102 @@
+import matplotlib.pyplot as plt
+import matplotlib as cm
+from matplotlib.colors import ListedColormap, LinearSegmentedColormap
+import os
+import glob
+import IPython
+import numpy as np
+from IPython import embed
+from scipy.optimize import curve_fit
+from jar_functions import parse_dataset
+from jar_functions import parse_infodataset
+from jar_functions import mean_traces
+from jar_functions import mean_noise_cut_eigen
+from jar_functions import norm_function
+from jar_functions import step_response
+from jar_functions import sort_values
+from jar_functions import average
+
+base_path = 'D:\\jar_project\\JAR\\eigen\\step'
+
+identifier = ['step_2015eigen8',
+              'step_2015eigen15',
+              'step_2015eigen16',
+              'step_2015eigen17',
+              'step_2015eigen19']
+datasets = []
+
+#dat = glob.glob('D:\\jar_project\\JAR\\2020*\\beats-eod.dat')
+#infodat = glob.glob('D:\\jar_project\\JAR\\2020*\\info.dat')
+
+time_all = []
+freq_all = []
+
+ID = []
+col = ['dimgrey', 'grey', 'darkgrey', 'silver', 'lightgrey', 'gainsboro', 'whitesmoke']
+labels = zip(ID, datasets)
+
+for infodataset in datasets:
+    infodataset = os.path.join(base_path, infodataset, 'info.dat')
+    i = parse_infodataset(infodataset)
+    identifier = i[0]
+    ID.append(identifier)
+
+for ID in identifier:
+    base_path = 'D:\\jar_project\\JAR\\eigenmannia\\step\\%s' %ID
+    res_df = []
+    for idx, dataset in enumerate(os.listdir(base_path)):
+        dataset = os.path.join(base_path, dataset, 'beats-eod.dat')
+        print(dataset)
+        #input of the function
+        frequency, time, amplitude, eodf, deltaf, stimulusf, duration, pause = parse_dataset(dataset)
+        dm = np.mean(duration)
+        pm = np.mean(pause)
+        timespan = dm + pm
+        start = np.mean([t[0] for t in time])
+        stop = np.mean([t[-1] for t in time])
+        if len(frequency) == 5:
+            continue
+
+        norm, base, jar = norm_function(frequency, time, onset_point=dm - dm, offset_point=dm)  # dm-dm funktioniert nur wenn onset = 0 sec
+        print(jar)
+        if jar[0] == 0.0:
+            continue
+
+        mf, tnew = mean_traces(start, stop, timespan, norm, time) # maybe fixed timespan/sampling rate
+
+        cf, ct = mean_noise_cut_eigen(mf, tnew, n=1250)
+
+        cf_arr = np.array(cf)
+        ct_arr = np.array(ct)
+
+        freq_all.append(cf_arr)
+        time_all.append(ct_arr)
+
+        b = []
+        for index, i in enumerate(ct):
+            if i > -45 and i < -5:
+                b.append(cf[index])
+
+        j = []
+        for indexx, h in enumerate(ct):
+            if h > 195 and h < 145:
+                j.append(cf[indexx])
+                print(h)
+                print(indexx)
+                print(cf[indexx])
+        r = np.median(j) - np.median(b)
+        #response.append(r)
+        embed()
+    base_line = plt.axhline(y = 0, color = 'black', ls = 'dotted', linewidth = '1')
+
+    plt.xlim([-10,220])
+    plt.xlabel('time [s]')
+    plt.ylabel('rel. JAR magnitude')
+    plt.title('relative JAR')
+    plt.savefig('relative JAR')
+    plt.legend(loc = 'lower right')
+    plt.show()
+    embed()
+
+
+# natalie fragen ob sie bei verschiedenen Amplituden messen kann (siehe tim)

notes

@@ -2,12 +2,16 @@
 eod basefrequenz rausziehen, scatter plot gegen cutoff frequency, ...
     - cutoff - dominance score
     - cutoff - basefrequency
-    - gain - dominance_score: für gain predict machen pro fish?
+    - gain - dominance_score: für gain predict machen pro fish,
+    hab ich dazu die richtige zeitckonstante aus gain_fit?
+    ... da ich ja prediction auch über sin und nicht step mache dann
 + eigenmannia: specgram von pre_data neben specgram von data machen um zu sehen ob analyse fehler oder fehler in import_data
     - erkenntnis: hab bei bm/jm nicht den gleichen mean abgezogen..
     - an sich res_df besser, jedoch immer noch relativ variabel
     - -2Hz bei meisten negative JAR?
     - evtl. doch mean anstatt median für response am ende?
++ look at step eigen data
+    - norming of data: what if in norm = ground / jar with jar == 0.0?
 + look at 5Hz data - compare
 
 long term: