Note

Click here to download the full example code

07.c stats.2dbin with fake data

Use binned_statistic_2d and display using heatmap.

C-Reactive Protein, count, median

Out:

[1 1 1 1 2 2 2 3]
[1 1 2 2 3 3 4 1]
[1 1 5 6 7 8 7 7]

Scipy:
[0.5 1.5 2.5 3.5 4.5 5.5 6.5]
[1.         1.03030303 1.06060606 1.09090909 1.12121212 1.15151515
 1.18181818 1.21212121 1.24242424 1.27272727 1.3030303  1.33333333
 1.36363636 1.39393939 1.42424242 1.45454545 1.48484848 1.51515152
 1.54545455 1.57575758 1.60606061 1.63636364 1.66666667 1.6969697
 1.72727273 1.75757576 1.78787879 1.81818182 1.84848485 1.87878788
 1.90909091 1.93939394 1.96969697 2.         2.03030303 2.06060606
 2.09090909 2.12121212 2.15151515 2.18181818 2.21212121 2.24242424
 2.27272727 2.3030303  2.33333333 2.36363636 2.39393939 2.42424242
 2.45454545 2.48484848 2.51515152 2.54545455 2.57575758 2.60606061
 2.63636364 2.66666667 2.6969697  2.72727273 2.75757576 2.78787879
 2.81818182 2.84848485 2.87878788 2.90909091 2.93939394 2.96969697
 3.         3.03030303 3.06060606 3.09090909 3.12121212 3.15151515
 3.18181818 3.21212121 3.24242424 3.27272727 3.3030303  3.33333333
 3.36363636 3.39393939 3.42424242 3.45454545 3.48484848 3.51515152
 3.54545455 3.57575758 3.60606061 3.63636364 3.66666667 3.6969697
 3.72727273 3.75757576 3.78787879 3.81818182 3.84848485 3.87878788
 3.90909091 3.93939394 3.96969697 4.        ]
[[2. 0. 1. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [2. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 2. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0. 0.]
 [0. 1. 0. 0. 0. 0.]]
[1.01515152 1.04545455 1.07575758 1.10606061 1.13636364 1.16666667
 1.1969697  1.22727273 1.25757576 1.28787879 1.31818182 1.34848485
 1.37878788 1.40909091 1.43939394 1.46969697 1.5        1.53030303
 1.56060606 1.59090909 1.62121212 1.65151515 1.68181818 1.71212121
 1.74242424 1.77272727 1.8030303  1.83333333 1.86363636 1.89393939
 1.92424242 1.95454545 1.98484848 2.01515152 2.04545455 2.07575758
 2.10606061 2.13636364 2.16666667 2.1969697  2.22727273 2.25757576
 2.28787879 2.31818182 2.34848485 2.37878788 2.40909091 2.43939394
 2.46969697 2.5        2.53030303 2.56060606 2.59090909 2.62121212
 2.65151515 2.68181818 2.71212121 2.74242424 2.77272727 2.8030303
 2.83333333 2.86363636 2.89393939 2.92424242 2.95454545 2.98484848
 3.01515152 3.04545455 3.07575758 3.10606061 3.13636364 3.16666667
 3.1969697  3.22727273 3.25757576 3.28787879 3.31818182 3.34848485
 3.37878788 3.40909091 3.43939394 3.46969697 3.5        3.53030303
 3.56060606 3.59090909 3.62121212 3.65151515 3.68181818 3.71212121
 3.74242424 3.77272727 3.8030303  3.83333333 3.86363636 3.89393939
 3.92424242 3.95454545 3.98484848]
[1. 2. 3. 4. 5. 6.]

'\nimport sys\nsys.exit()\n# Compute bin statistic (count and median)\n\n# Plot\nplt.figure()\nsns.violinplot(data=data, x="timestep", y="shap_values", inner="box")\nplt.figure()\nplt.tight_layout()\nsns.violinplot(data=data, x="timestep", y="feature_values", inner="box")\nplt.figure()\nsns.histplot(data=data, x="timestep", shrink=.8)\n\n\n# Plot hist\nf1 = plt.hist2d(data.timestep, data.feature_values, bins=30, cmap=\'Reds\')\ncb = plt.colorbar()\ncb.set_label(\'counts in bin\')\nplt.title(\'Counts (square bin)\')\nplt.show()\n'


  9 # Libraries
 10 import seaborn as sns
 11 import pandas as pd
 12 import numpy as np
 13 import matplotlib.pyplot as plt
 14
 15 from scipy import stats
 16 from matplotlib.colors import LogNorm
 17
 18 def data_shap():
 19     data = pd.read_csv('../../datasets/shap/shap.csv')
 20     return data.timestep, \
 21            data.shap_values, \
 22            data.feature_values, \
 23            data
 24
 25 def data_manual():
 26     """"""
 27     # Create random values
 28     x = np.array([1, 1, 1, 1, 2, 2, 2, 3])
 29     y = np.array([1, 1, 2, 2, 3, 3, 4, 1])
 30     z = np.array([1, 1, 5, 6, 7, 8, 7, 7])
 31     return x, y, z, None
 32
 33 # Create data
 34 x, y, z, data = data_manual()
 35
 36 print(x)
 37 print(y)
 38 print(z)
 39
 40 """
 41 # With pandas
 42 v = z
 43 vals, bins = np.histogram(v)
 44 a = pd.Series(v).groupby(pd.cut(v, bins)).median()
 45 print("\nPandas:")
 46 print(bins)
 47 print(vals)
 48 print(a)
 49 """
 50
 51
 52
 53
 54
 55
 56
 57 vmin = z.min()
 58 vmax = z.max()
 59
 60
 61
 62 # Compute binned statistic (median)
 63 binx = np.linspace(0, 3, 4) + 0.5
 64 biny = np.linspace(0, 4, 5) + 0.5
 65 binx = np.linspace(0, 6, 7) + 0.5
 66 biny = np.linspace(y.min(), y.max(), 100)
 67 r1 = stats.binned_statistic_2d(x=y, y=x, values=z,
 68     statistic='count', bins=[biny, binx],
 69     expand_binnumbers=False)
 70
 71 r2 = stats.binned_statistic_2d(x=y, y=x, values=z,
 72     statistic='median', bins=[biny, binx],
 73     expand_binnumbers=False)
 74
 75 # Compute centres
 76 x_center = (r1.x_edge[:-1] + r1.x_edge[1:]) / 2
 77 y_center = (r1.y_edge[:-1] + r1.y_edge[1:]) / 2
 78
 79 # Show
 80 print("\nScipy:")
 81 print(binx)
 82 print(biny)
 83 print(r1.statistic)
 84 print(x_center)
 85 print(y_center)
 86
 87 # Convert the computed matrix to an stacked dataframe?
 88
 89 flip1 = np.flip(r1.statistic, 0)
 90 flip2 = np.flip(r2.statistic, 0)
 91 #flip1 = r1.statistic
 92 #flip2 = r2.statistic
 93
 94
 95
 96 # Display
 97 fig, axs = plt.subplots(nrows=1, ncols=2,
 98     sharey=False, sharex=False, figsize=(8, 7))
 99
100 sns.heatmap(flip1, annot=False, linewidth=0.5,
101     xticklabels=y_center.astype(int),
102     yticklabels=x_center.round(2)[::-1], # Because of flip
103     cmap='Blues', ax=axs[0],
104     norm=LogNorm())
105
106 sns.heatmap(flip2, annot=False, linewidth=0.5,
107     xticklabels=y_center.astype(int),
108     yticklabels=x_center.round(2)[::-1], # Because of flip
109     cmap='coolwarm', ax=axs[1], zorder=1,
110     vmin=None, vmax=None, center=None, robust=True)
111
112 # If robust=True and vmin or vmax are absent, the colormap range
113 # is computed with robust quantiles instead of the extreme values.
114 """
115 sns.violinplot(x=x, y=y,
116     saturation=0.5, fliersize=0.1, linewidth=0.5,
117     color='green', ax=axs[2], zorder=3,
118     width=0.5)
119 """
120
121
122 # Configure ax0
123 axs[0].set_title('count')
124 axs[0].set_xlabel('timestep')
125 axs[0].set_ylabel('shap')
126 axs[0].locator_params(axis='y', nbins=10)
127 #axs[0].set_aspect('equal', 'box'
128
129 # Configure ax1
130 axs[1].set_title('median')
131 axs[1].set_xlabel('timestep')
132 axs[1].set_ylabel('shap')
133 axs[1].locator_params(axis='y', nbins=10)
134 #axs[1].set_aspect('equal', 'box')
135 #axs[1].invert_yaxis()
136
137 # Generic
138 plt.suptitle('C-Reactive Protein')
139
140 """
141 # Set axes manually
142 #plt.set_xticks()
143 #plt.setp(axs[1].get_yticklabels()[::1], visible=False)
144 #plt.setp(axs[1].get_yticklabels()[::5], visible=True)
145 from matplotlib import ticker
146 axs[1].xaxis.set_major_locator(ticker.MultipleLocator(1.00))
147 axs[1].xaxis.set_minor_locator(ticker.MultipleLocator(0.25))
148 """
149
150 plt.tight_layout()
151 plt.show()
152
153
154 """
155 import sys
156 sys.exit()
157 # Compute bin statistic (count and median)
158
159 # Plot
160 plt.figure()
161 sns.violinplot(data=data, x="timestep", y="shap_values", inner="box")
162 plt.figure()
163 plt.tight_layout()
164 sns.violinplot(data=data, x="timestep", y="feature_values", inner="box")
165 plt.figure()
166 sns.histplot(data=data, x="timestep", shrink=.8)
167
168
169 # Plot hist
170 f1 = plt.hist2d(data.timestep, data.feature_values, bins=30, cmap='Reds')
171 cb = plt.colorbar()
172 cb.set_label('counts in bin')
173 plt.title('Counts (square bin)')
174 plt.show()
175 """

Total running time of the script: ( 0 minutes 0.853 seconds)

Gallery generated by Sphinx-Gallery