imports - 3

examples/classic_fit_example.py

@@ -19,17 +19,15 @@
     for c in range(int(charge)):
         data.append(time)
 
-
 def gauss(x, *p):
     '''
     The gaussian distribution.
     '''
     amplitude, mean, sigma = p
-    return amplitude * numpy.exp(-(x - mean) ** 2 / (2. * sigma ** 2))
-
+    return amplitude*numpy.exp(-(x-mean)**2/(2.*sigma**2))
 
 hist, bin_edges = numpy.histogram(data)
-bin_centers = (bin_edges[:-1] + bin_edges[1:]) / 2
+bin_centers = (bin_edges[:-1] + bin_edges[1:])/2
 p0 = [1000., 0., 1.]
 p1 = [1000., 0., 1.]
 print(hist)

examples/concept_illustration.py

@@ -11,9 +11,8 @@
 '''
 
 import numpy
-import pylab  # type: ignore[import]
+import pylab   # type: ignore[import]
 
-import voka.model
 import voka.tools.render
 import voka.two_sample
 
@@ -32,10 +31,10 @@
     histograms['ChargeHistogram%d' % i] = numpy.histogram(charge_data, bins=100)
 
 # Render the histograms
-i = 0
+i=0
 for title, histogram in histograms.items():
     pylab.figure(i)
-    i = i + 1
+    i=i+1
     bin_values = histogram[0]
     voka.tools.render.draw(bin_values, title)
 
@@ -45,7 +44,7 @@
 # First compare the histograms with traditional statistical tests
 test_sample = histograms['ChargeHistogram0'][0]
 benchmark_ensemble = [histograms['ChargeHistogram%d' % i][0]
-                      for i in range(1, 6)]
+                      for i in range(1,6)]
 print(benchmark_ensemble)
 for benchmark_sample in benchmark_ensemble:
     result = voka.two_sample.traditional(test_sample, benchmark_sample)
@@ -54,8 +53,8 @@
 # Second compare the test histogram using voka.
 model = voka.model.Voka()
 reference_collection = {
-    "Benchmark%d" % idx:
-        {"ChargeHistogram": s}
+    "Benchmark%d" % idx :
+    {"ChargeHistogram":s}
     for idx, s in enumerate(benchmark_ensemble)
 }
 

examples/standard_distribution_comparisons.py

@@ -14,6 +14,7 @@
 
 import numpy
 import pylab  # type: ignore[import]
+
 # What's the difference here between voka.metrics.chisq
 # and scipy.stats.chisquare ?
 # import voka.metrics.chisq
@@ -34,42 +35,42 @@
     data = numpy.random.normal(size=SIZE)
     histograms['Gaussian%d' % i] = numpy.histogram(data, bins=BINS)
 
-    data = numpy.random.normal(size=int(SIZE / 2), loc=-1, scale=0.25) + \
-           numpy.random.normal(size=int(SIZE / 2), loc=1, scale=0.25)
+    data = numpy.random.normal(size=int(SIZE/2), loc=-1, scale=0.25) + \
+        numpy.random.normal(size=int(SIZE/2), loc=1, scale=0.25)
     histograms['BiGaussian%d' % i] = numpy.histogram(data, bins=BINS)
 
     data = numpy.random.exponential(scale=10, size=1000)
     histograms['Exponential%d' % i] = numpy.histogram(data, bins=BINS)
 
 T_dist = collections.defaultdict(list)
-# test_stat = voka.metrics.chisq.NormChiSq()
+#test_stat = voka.metrics.chisq.NormChiSq()
 test_stat = scipy.stats.chisquare
 print("Comparing histograms...")
 for i in range(N_HISTOGRAMS):
-    for j in range(i + 1, N_HISTOGRAMS):
+    for j in range(i+1, N_HISTOGRAMS):
         h1 = histograms["Uniform%d" % i][0]
         h2 = histograms["Uniform%d" % j][0]
-        # T_dist['uniform'].append(test_stat(h1, h2))
+        #T_dist['uniform'].append(test_stat(h1, h2))
         T_dist['uniform'].append(test_stat(h1, h2)[0])
 
         h1 = histograms["ChiSq5%d" % i][0]
         h2 = histograms["ChiSq5%d" % j][0]
-        # T_dist['chisq'].append(test_stat(h1, h2))
+        #T_dist['chisq'].append(test_stat(h1, h2))
         T_dist['chisq'].append(test_stat(h1, h2)[0])
 
         h1 = histograms["Exponential%d" % i][0]
         h2 = histograms["Exponential%d" % j][0]
-        # T_dist['exponential'].append(test_stat(h1, h2))
+        #T_dist['exponential'].append(test_stat(h1, h2))
         T_dist['exponential'].append(test_stat(h1, h2)[0])
 
         h1 = histograms["Gaussian%d" % i][0]
         h2 = histograms["Gaussian%d" % j][0]
-        # T_dist['gaussian'].append(test_stat(h1, h2))
+        #T_dist['gaussian'].append(test_stat(h1, h2))
         T_dist['gaussian'].append(test_stat(h1, h2)[0])
 
         h1 = histograms["BiGaussian%d" % i][0]
         h2 = histograms["BiGaussian%d" % j][0]
-        # T_dist['bigaussian'].append(test_stat(h1, h2))
+        #T_dist['bigaussian'].append(test_stat(h1, h2))
         T_dist['bigaussian'].append(test_stat(h1, h2)[0])
 
 pylab.figure(1)

examples/stochastic_example.py

@@ -35,7 +35,7 @@
 
 NDOF = None
 for i in range(N_HISTOGRAMS):
-    for j in range(i + 1, N_HISTOGRAMS):
+    for j in range(i+1, N_HISTOGRAMS):
         ch1 = histograms["ChargeHistogram%d" % i][0]
         ch2 = histograms["ChargeHistogram%d" % j][0]
         th1 = histograms["TimeHistogram%d" % i][0]

examples/traditional_two_sample_methods.py

@@ -23,7 +23,6 @@
 
 import voka.tools.render
 
-
 # Low p-value is bad here.
 # The null hypothesis H_0 is that the two samples
 # were sampled from the same underlying distribution.
@@ -116,14 +115,14 @@ def voka_2sample(sample1, sample2):
     except:
         print("    skipping kruskal")
 
-    #    try:
-    #        r = scipy.stats.friedmanchisquare(sample1, sample2)
-    #        result['FriedmanChiSquare'] = {
-    #            'statistic': r.statistic,
-    #            'pvalue': r.pvalue
-    #        }
-    #    except ValueError:
-    #        print("    skipping friedmanchisquare")
+#    try:
+#        r = scipy.stats.friedmanchisquare(sample1, sample2)
+#        result['FriedmanChiSquare'] = {
+#            'statistic': r.statistic,
+#            'pvalue': r.pvalue
+#        }
+#    except ValueError:
+#        print("    skipping friedmanchisquare")
 
     r = scipy.stats.brunnermunzel(sample1, sample2)
     result['BrunnerMunzel'] = {
@@ -133,13 +132,11 @@ def voka_2sample(sample1, sample2):
 
     return result
 
-
 DENSITY = True
 
-
 def compare(benchmark_distribution, systematic_distribution, pvalues, density=False):
     # Use the same histogram settings for each.
-    _range = (-5, 5)
+    _range = (-5,5)
     _bins = 100
 
     # First histogram to make samples
@@ -157,7 +154,6 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
     for k, v in comparison.items():
         pvalues[k].append(v['pvalue'])
 
-
 if __name__ == '__main__':
 
     # Tests to perform
@@ -170,32 +166,32 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
     benchmark_center = 0.0
     benchmark_size = 100000
 
-    widths = numpy.arange(0.5 * benchmark_width,
-                          1.5 * benchmark_width,
+    widths = numpy.arange(0.5*benchmark_width,
+                          1.5*benchmark_width,
                           0.001)
 
     centers = numpy.arange(-0.5,
                            0.5,
                            0.001)
 
-    sizes = numpy.arange(0.5 * benchmark_size,
-                         1.5 * benchmark_size,
+    sizes = numpy.arange(0.5*benchmark_size,
+                         1.5*benchmark_size,
                          10)
 
-    draw_width_idx = int(0.9 * len(widths))
-    draw_center_idx = int(0.9 * len(centers))
+    draw_width_idx = int(0.9*len(widths))
+    draw_center_idx = int(0.9*len(centers))
 
     benchmark_distribution = numpy.random.normal(loc=benchmark_center,
                                                  scale=benchmark_width,
                                                  size=benchmark_size)
 
     # Generate systematic distributions
-    print(80 * '-')
+    print(80*'-')
     figure_number = 1
 
-    pvalues_width_systematic: dict = collections.defaultdict(list)
+    pvalues_width_systematic = collections.defaultdict(list)
     for idx, systematic_width in enumerate(widths):
-        # print('width = %.4f' % systematic_width)
+        #print('width = %.4f' % systematic_width)
         systematic_distribution = numpy.random.normal(loc=benchmark_center,
                                                       scale=systematic_width,
                                                       size=benchmark_size)
@@ -207,7 +203,7 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
         if idx == draw_width_idx:
             print('width = %.4f' % systematic_width)
             # Use the same histogram settings for each.
-            _range = (-5, 5)
+            _range = (-5,5)
             _bins = 100
 
             # First histogram to make samples
@@ -224,6 +220,7 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
             figure_number += 1
             voka.tools.render.draw_ratio(systematic_sample, benchmark_sample)
 
+
     pylab.figure(figure_number)
     figure_number += 1
     for label, pvs in pvalues_width_systematic.items():
@@ -236,10 +233,10 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
     pylab.xlabel('width')
     pylab.ylabel('log(p-value)')
 
-    print(80 * '-')
-    pvalues_center_systematic: dict = collections.defaultdict(list)
+    print(80*'-')
+    pvalues_center_systematic = collections.defaultdict(list)
     for idx, systematic_center in enumerate(centers):
-        # print('center = %.4f' % systematic_center)
+        #print('center = %.4f' % systematic_center)
         systematic_distribution = numpy.random.normal(loc=systematic_center,
                                                       scale=benchmark_width,
                                                       size=benchmark_size)
@@ -251,7 +248,7 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
         if idx == draw_center_idx:
             print('center = %.4f' % systematic_center)
             # Use the same histogram settings for each.
-            _range = (-5, 5)
+            _range = (-5,5)
             _bins = 100
 
             # First histogram to make samples
@@ -279,12 +276,12 @@ def compare(benchmark_distribution, systematic_distribution, pvalues, density=Fa
     pylab.xlabel('center')
     pylab.ylabel('log(p-value)')
 
-    print(80 * '-')
+    print(80*'-')
     pylab.figure(figure_number)
     figure_number += 1
-    pvalues_size_systematic: dict = collections.defaultdict(list)
+    pvalues_size_systematic = collections.defaultdict(list)
     for systematic_size in sizes:
-        # print('size = %d' % int(systematic_size))
+        #print('size = %d' % int(systematic_size))
         systematic_distribution = numpy.random.normal(loc=benchmark_center,
                                                       scale=benchmark_width,
                                                       size=int(systematic_size))