-
Notifications
You must be signed in to change notification settings - Fork 2
/
ash_testing_example.py
executable file
·114 lines (79 loc) · 3.26 KB
/
ash_testing_example.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
# -*- coding: utf-8 -*-
"""
Created on Fri Jul 24 13:09:50 2015
@author: bcolsen
"""
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import scipy.stats as stats
from ash import ash
plt.rcParams['svg.fonttype'] = 'none'
# %% Make Fake data and store it in excell.
# Don't do this if you have real data
mu, sigma = 6.35, 0.13
data_fake_a = mu + sigma*np.random.randn(50)
mu2, sigma2 = 6.5, 0.10
data_fake_b = mu2 + sigma2*np.random.randn(50)
df_fake = pd.DataFrame({'fake_a':data_fake_a, 'fake_b':data_fake_b})
df_fake.to_excel('fake_data.xlsx')
# %% Import tha data and assign a and b
filename = 'fake_data.xlsx' #Change this
df = pd.read_excel('fake_data.xlsx')
a = df['fake_a']
b = df['fake_b']
label_a = "Fake A"
label_b = "Fake B"
xlabel = 'Fakeness (%)'
fig = plt.figure(num = 'ASH Plot testing', figsize = (4,4))
fig.clf()
ash_obj_a = ash(a)
ash_obj_b = ash(b)
ax = plt.subplot()
#plot ASH as a line
ax.plot(ash_obj_b.ash_mesh,ash_obj_b.ash_den,lw=2, color = '#D95319')
ax.plot(ash_obj_a.ash_mesh,ash_obj_a.ash_den,lw=2, color = '#365994')
#plot the solid ASH
ash_obj_b.plot_ash_infill(ax, color ='#F2966E')
ash_obj_a.plot_ash_infill(ax, color='#92B2E7')
#plot KDE
ax.plot(ash_obj_a.kde_mesh,ash_obj_a.kde_den,lw=1, color ='#365994')
ax.plot(ash_obj_b.kde_mesh,ash_obj_b.kde_den,lw=1, color = '#D95319')
# Make a Rugplot (the barcode like data representation)
ash_obj_a.plot_rug(ax, alpha=1, color = '#4C72B0', ms = 8, height = 0.10)
ash_obj_b.plot_rug(ax, alpha=1, color ='#F2966E', ms = 8, height = 0.04)
if ash_obj_a.mean <= ash_obj_b.mean:
ash_obj_a.plot_stats(ax, label_a, color = '#365994')
ash_obj_b.plot_stats(ax, label_b, side='right', color ='#D95319')
else:
ash_obj_a.plot_stats(ax, label_a, side='right', color = '#365994')
ash_obj_b.plot_stats(ax, label_b, color ='#D95319')
# add a line showing the expected distribution
dista = stats.norm.pdf(ash_obj_a.ash_mesh, mu, sigma)
plt.plot(ash_obj_a.ash_mesh,dista,'b--',lw=2)
plt.plot(ash_obj_a.ash_mesh,dista-ash_obj_a.ash_den,'b--',lw=1)
distb = stats.norm.pdf(ash_obj_b.ash_mesh, mu2, sigma2)
plt.plot(ash_obj_b.ash_mesh,distb,'r--',lw=2)
plt.plot(ash_obj_b.ash_mesh,distb-ash_obj_b.ash_den,'r--',lw=1)
ax.yaxis.set_ticks_position('left')
ax.xaxis.set_ticks_position('bottom')
ax.tick_params(direction='out')
ax.set_yticks([])
print(label_a, ash_obj_a.mean, ash_obj_a.sigma)
ad_stat, ad_crit, ad_signi = stats.anderson_ksamp((a,b))
print('ad',label_a, ad_stat, ad_crit, ad_signi)
ks_stat, ks_signi = stats.ks_2samp(a,b)
print('ks',label_a, ks_stat, ks_signi)
print(ash_obj_a.bw, ash_obj_a.bin_num, ash_obj_a.bin_width, ash_obj_a.bin_edges[1]-ash_obj_a.bin_edges[0])
print(ash_obj_b.bw, ash_obj_b.bin_num, ash_obj_b.bin_width, ash_obj_b.bin_edges[1]-ash_obj_b.bin_edges[0])
ad_ks_str = "p = {:.2f} %".format(ad_signi*100, ks_signi*100)
#p value text
ax.text(0.05, 0.75, ad_ks_str, color='k', ha='left', va='center', transform=ax.transAxes, size=14)
plt.xlabel(xlabel)
plt.tight_layout()
plt.subplots_adjust(top=0.95)
fig.text(0.47, 0.97, label_a, size=14, color='#365994', ha='right')
fig.text(0.5, 0.97, 'vs.', size=14, ha='center')
fig.text(0.53, 0.97, label_b, size=14, color='#D95319', ha='left')
fig.savefig(label_a + '_vs_' + label_b + 'testing.svg', dpi=300, transparent=True)
plt.show()