view experiment-reverb/code/plots.py @ 2:c87a9505f294 tip

Added LICENSE for code, removed .wav files

# -*- coding: utf-8 -*-
"""
Created on Mon Jul 13 06:45:38 2015

@author: Emmanouil Theofanis Chourdakis
"""


# Plots for our paper

from numpy import *
import matplotlib.pyplot as plt
from matplotlib import rc
#from matplotlib2tikz import save as tikz_save
rc('font',**{'family':'serif','serif':['Palatino']})
rc('text', usetex=True)
rc('pgf', texsystem='pdflatex')

sucrats1 = (0.716, 0.793, 0.703, 0.696, 0.820)
sucrats2 = (0.789, 0.789, 0.729, 0.525, 0.793)
sucrats3 = (0.769, 0.736, 0.760, 0.537, 0.709)
sucrats4 = (0.689, 0.689, 0.552, 0.562, 0.673)
sucrats5 = (0.80508403361344549, 0.77222689075630258, 0.76915966386554613, 0.5345168067226892, 0.7770588235294118)
sucrats6 = (0.796, 0.898, 0.798, 0.563, 0.875)

example = {'SIM/C1':(0.5,0.7,0.6,0.8), 'FUL/C1':(0.6,0.7,0.9,0.8), 'SIM/C2': (0.4,0.3,0.5,0.6)}

FONTSIZE=20

GNBCR = []
SVMR  = []
HMMCR = []
HMMSVMR = []
SINKHOLE = []




labels = ['Set 1', 'Set 2', 'Set 3', 'Set 4', 'Set 5', 'Set 6']

results = [sucrats1, sucrats2, sucrats3, sucrats4]

for i in range(0, len(results)):
    GNBCR.append(results[i][0])
    SVMR.append(results[i][1])
    HMMCR.append(results[i][2])
    HMMSVMR.append(results[i][3])
    SINKHOLE.append(results[i][4])
    

pos = list(range(len(SVMR)))
width = 0.15

fig,ax=plt.subplots(figsize=(10,10))

bar1=plt.bar(pos, GNBCR, width,
                 alpha=0.5,
                 color='r',
                 hatch='x', # this one defines the fill pattern
                 label=labels[0])

plt.bar([p + width for p in pos], SVMR, width,
                 alpha=0.5,
                 color='g',
                 hatch='-',
                 label=labels[1])
    
plt.bar([p + width*2 for p in pos], HMMCR, width,
                 alpha=0.5,
                 color='b',
                 hatch='',
                 label=labels[2])

plt.bar([p + width*3 for p in pos], HMMSVMR, width,
                 alpha=0.5,
                 color='c',hatch='/',
                 label=labels[3])

plt.bar([p + width*4 for p in pos], SINKHOLE, width,
                 alpha=0.5,
                 color='m',hatch='\\',
                 label=labels[3])                 
                 


# Setting axis labels and ticks
ax.set_ylabel('Success Ratio', fontsize=FONTSIZE)
ax.set_xlabel('Data set', fontsize=FONTSIZE)
ax.set_title('Classifier Success Ratio', fontsize=FONTSIZE)
ax.set_xticks([p + 2 * width for p in pos])
ax.set_yticks([0, 0.2, 0.6, 0.8, 1.0])
ax.set_xticklabels(labels,  fontsize=FONTSIZE)
ax.set_yticklabels([0, 0.2, 0.6, 0.8, 1.0], fontsize=FONTSIZE)

# Setting the x-axis and y-axis limits
plt.xlim(min(pos)-width, max(pos)+width*6)
plt.ylim([0,1.2])

# Adding the legend and showing the plot
leg = plt.legend(['GNB', 'SVM', 'HMM', 'HMM/SVM', 'SINK-HOLE'], loc='upper right', fontsize=FONTSIZE, fancybox=True)

# leg.get_frame().set_alpha(0.5)
plt.grid()
#plt.show()

#tikz_save('plot.tkz', figureheight='4cm', figurewidth='6cm')
fig.tight_layout()
fig.savefig('./plot.pgf', dpi=500)

from sklearn import metrics
def plot_confusion_matrix(y_pred, y):
    plt.imshow(metrics.confusion_matrix(y, y_pred),
               cmap=plt.cm.binary, interpolation='nearest')
    plt.colorbar()
    plt.xlabel('true value')
    plt.ylabel('predicted value')
    
#plt.figure()    
#plot_confusion_matrix(predhmmc3, parameters_state)


msecrats1 = (0.015, 0.013, 0.019, 0.012, 0.007)
msecrats2 = (0.005, 0.006, 0.009, 0.007, 0.004)
msecrats3 = (0.018, 0.020, 0.014, 0.019, 0.019)
msecrats4 = (0.010, 0.010, 0.018, 0.010, 0.010)
msecrats5 = (0.097, 0.014, 0.013, 0.017, 0.010)
msecrats6 = (0.006, 0.003, 0.012, 0.013, 0.003)


results = [msecrats1, msecrats2, msecrats3, msecrats4, msecrats5, msecrats6]

GNBCR = []
SVMR  = []
HMMCR = []
HMMSVMR = []
SINKHOLE = []
for i in range(0, len(results)):
    GNBCR.append(results[i][0])
    SVMR.append(results[i][1])
    HMMCR.append(results[i][2])
    HMMSVMR.append(results[i][3])
    SINKHOLE.append(results[i][4])
    
pos = list(range(len(SVMR)))

plt.close('all')
#plt.figure()
fig,ax=plt.subplots(figsize=(10,10))



bar1=plt.barh(pos, GNBCR, width,
                 alpha=0.5,
                 color='r',
                 hatch='x', # this one defines the fill pattern
                 label=labels[0])

plt.barh([p + width for p in pos], SVMR, width,
                 alpha=0.5,
                 color='g',
                 hatch='-',
                 label=labels[1])
    
plt.barh([p + width*2 for p in pos], HMMCR, width,
                 alpha=0.5,
                 color='b',
                 hatch='',
                 label=labels[2])

plt.barh([p + width*3 for p in pos], HMMSVMR, width,
                 alpha=0.5,
                 color='c',hatch='/',
                 label=labels[3])

plt.barh([p + width*4 for p in pos], SINKHOLE, width,
                 alpha=0.5,
                 color='m',hatch='\\',
                 label=labels[3])      


# Setting axis labels and ticks
ax.set_ylabel('Data set', fontsize=FONTSIZE)
ax.set_xlabel('Mean Squared Error', fontsize=FONTSIZE)
ax.set_title('Mean Squared Errors', fontsize=FONTSIZE)
ax.set_yticks([p + 2 * width for p in pos])
ax.set_xticks([0, 0.02])
ax.set_yticklabels(labels,  fontsize=FONTSIZE)
ax.set_xticklabels([0, 0.02], fontsize=FONTSIZE)

# Setting the x-axis and y-axis limits
plt.ylim(min(pos)-width, max(pos)+width*6)
plt.xlim([0,0.03])

# Adding the legend and showing the plot
leg = plt.legend(['GNB', 'SVM', 'HMM', 'HMM/SVM', 'SINK-HOLE'], loc='upper right', fontsize=FONTSIZE, fancybox=True)

# leg.get_frame().set_alpha(0.5)
plt.grid()


#tikz_save('plot.tkz', figureheight='4cm', figurewidth='6cm')
fig.tight_layout()
fig.savefig('./plotmses.pgf', dpi=500)

from sklearn import metrics
def plot_confusion_matrix(y_pred, y):
    plt.imshow(metrics.confusion_matrix(y, y_pred),
               cmap=plt.cm.binary, interpolation='nearest')
    plt.colorbar()
    plt.xlabel('true value')
    plt.ylabel('predicted value')
#plt.show()