Functions/PC_traj.py at master · ilbarlow/Functions · GitHub

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
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import matplotlib.animation as animation
import numpy as np
import os
import re

def PC_trajGraded(PCmean, PCsem, PCsToPlot, rep, saveDir, file_type, scaling, start_end, cum_var, legend):

    """ New figure style to plot the PC trajectories with a gradated colorscheme according to scaling
     Input:
        PCmean - dataframe of the PCmeans calculated after doing PCA and transforming into PC space

        PCsem - sem of PCs

        PCsToPlot = principal components to plot

        rep - if replicated specify

        saveDir - directory into which the figure should be saved

        file_type - filetype eg .png, .svg, .eps

        scaling - how to scale and grade the data

        start_end - whether to write on start and end of trajectory

        cum_var - array of cumulative variance explained by each PC

        legend - 'on' or 'off'

        animation - whether to animate the plot (Boolean)

    Output:
    Figure

    """
    PC1 = PCsToPlot[0]
    PC2 = PCsToPlot[1]

    xscale = 1/(PCmean.max()[PC1] - PCmean.min()[PC1])
    yscale = 1/(PCmean.max()[PC2] - PCmean.min()[PC2])

    allDrugs= np.unique(PCmean['drug'])

    plt.figure()
    for drug in allDrugs:
        cscale = np.arange(1, np.unique(PCmean[PCmean['drug']==drug][scaling]).shape[0]+1,1)

        plt.errorbar(x= PCmean[PCmean['drug']==drug][PC1]*xscale,\
                     y = PCmean[PCmean['drug']==drug][PC2]*yscale,\
                     xerr = PCsem[PCsem['drug']==drug][PC1]*xscale,\
                     yerr = PCsem[PCsem['drug']==drug][PC2]*yscale, \
                     color = [0.8, 0.8, 0.8], zorder = -0.5, elinewidth = 2, label = None)
        plt.pause(0.1)
        plt.scatter(x = PCmean[PCmean['drug']==drug][PC1]*xscale,\
                    y=PCmean[PCmean['drug']==drug][PC2]*yscale,\
                    cmap = plt.get_cmap(drug),c=cscale , vmin = -1,\
                     vmax = cscale[-1]+3, label = drug, alpha =0.9)
    plt.axis('scaled')
    plt.xlim (-1,1)
    plt.ylim (-1,1)
    plt.title(scaling)
    try:
        plt.xlabel(PC1 + '_' + str(cum_var[0]*100) + '%')
        plt.ylabel(PC2 + '_'+ str((cum_var[1] - cum_var[0])*100) + '%')
    except TypeError:
        print ('no cumulate variance input')
    if legend == 'on':
        plt.legend(loc='upper left', bbox_to_anchor=(1.1,1.05) ,ncol = 1, frameon= True)

    plt.tight_layout()
    plt.savefig(os.path.join(saveDir, '{}_{}_PC_errorbar{}'.format(rep, PCsToPlot, file_type)), dpi = 200)

    plt.show()

  #   if animation:
  #   	Writer = animation.writers['ffmpeg']
		# writer = Writer(fps=20, metadata=dict(artist='Me'), bitrate=1800)


def PC_trajGradedNoScale(PCmean, PCsem, rep, saveDir, file_type, scaling, start_end, cum_var, legend):

    """ New figure style to plot the PC trajectories with a gradated colorscheme according to scaling
     Input:
        PCmean - dataframe of the PCmeans calculated after doing PCA and transforming into PC space

        PCsem - sem of PCs

        rep - if replicated specify

        saveDir - directory into which the figure should be saved

        file_type - filetype eg .png, .svg, .eps

        scaling - how to scale and grade the data

        start_end - whether to write on start and end of trajectory

        cum_var - array of cumulative variance explained by each PC

        legend - 'on' or 'off'

    Output:
    Figure

    """


    allDrugs= np.unique(PCmean['drug'])

    plt.figure()
    for drug in allDrugs:
        cscale = np.arange(1, np.unique(PCmean[PCmean['drug']==drug][scaling]).shape[0]+1,1)

        plt.errorbar(x= PCmean[PCmean['drug']==drug]['PC_1'],\
                     y = PCmean[PCmean['drug']==drug]['PC_2'],\
                     xerr = PCsem[PCsem['drug']==drug]['PC_1'],\
                     yerr = PCsem[PCsem['drug']==drug]['PC_2'], \
                     color = [0.8, 0.8, 0.8], zorder = -0.5, elinewidth = 2, label = None)
        plt.pause(0.1)
        plt.scatter(x = PCmean[PCmean['drug']==drug]['PC_1'],\
                    y=PCmean[PCmean['drug']==drug]['PC_2'],\
                    cmap = plt.get_cmap(drug),c=cscale , vmin = -1,\
                     vmax = cscale[-1]+3, label = drug, alpha =0.9)
    plt.axis('scaled')
    plt.title(scaling)
    plt.xlabel('PC_1 ' + str(cum_var[0]*100) + '%')
    plt.ylabel('PC_2 ' + str((cum_var[1] - cum_var[0])*100) + '%')
    if legend == 'on':
        plt.legend(loc='upper left', bbox_to_anchor=(1.1,1.05) ,ncol = 1, frameon= True)

    plt.tight_layout()
    plt.savefig(os.path.join(saveDir, str(rep) + 'PC12_errorbarNS' + file_type), dpi = 200)

    plt.show()