notebook.community

Edit and run



In [1]:

    
import pandas as pd
import numpy as np
%matplotlib inline
import seaborn as sns
import matplotlib.pyplot as plt
sns.set()
pd.set_option('display.precision', 4)
np.set_printoptions(precision=3)



In [12]:

    
raw_scores = pd.read_csv('../results_olda_reduced.csv', index_col='track_id')



In [13]:

    
del raw_scores['Unnamed: 0']



In [14]:

    
scores = raw_scores.dropna()



In [15]:

    
scores.head(5)









    Out[15]:






  
    
      
      t_measure10
      t_measure15
      t_measure30
      t_precision10
      t_precision15
      t_precision30
      t_recall10
      t_recall15
      t_recall30
    
    
      track_id
      
      
      
      
      
      
      
      
      
    
  
  
    
      SALAMI_10
      0.337
      0.354
      0.213
      0.242
      0.244
      0.126
      0.558
      0.644
      0.698
    
    
      SALAMI_100
      0.387
      0.503
      0.520
      0.317
      0.438
      0.407
      0.496
      0.592
      0.721
    
    
      SALAMI_1000
      0.374
      0.377
      0.394
      0.331
      0.306
      0.290
      0.430
      0.491
      0.611
    
    
      SALAMI_1002
      0.430
      0.418
      0.054
      0.361
      0.317
      0.028
      0.531
      0.614
      0.689
    
    
      SALAMI_1004
      0.192
      0.162
      0.120
      0.119
      0.093
      0.065
      0.498
      0.640
      0.791



In [16]:

    
scores.columns = ['T_F @ 10','T_F @ 15', 'T_F @ 30', 'T_P @ 10', 'T_P @ 15', 'T_P @ 30', 'T_R @ 10', 'T_R @ 15', 'T_R @ 30']



In [17]:

    
scores = scores[['T_F @ 10',  'T_P @ 10', 'T_R @ 10',
                 'T_F @ 15',  'T_P @ 15', 'T_R @ 15',
                 'T_F @ 30', 'T_P @ 30', 'T_R @ 30']]



In [18]:

    
scores.describe()



In [21]:

    
sns.set_style('darkgrid')
plt.figure(figsize=(6,3))
sns.boxplot(scores[['T_F @ 10', 'T_F @ 15', 'T_F @ 30']], color=sns.color_palette('Greens')[3])
plt.ylabel('$\mathcal{T}_F$')
plt.xticks([1,2,3], [10, 15, 30])
plt.ylim([0, 1])
plt.xlabel('$w$ (seconds)')
plt.tight_layout()
plt.savefig('../paper/figs/salami-olda-stats.pdf', pad_inches=0.0, bbox_inches='tight')
!convert ../paper/figs/salami-olda-stats.pdf ../paper/figs/salami-olda-stats.eps



In [20]:

    
sns.set_style('darkgrid')
plt.figure(figsize=(6,3))
#sns.boxplot(scores[['T_F @ 10', 'T_F @ 15', 'T_F @ 30']], color=sns.color_palette('Greens')[3])
sns.boxplot(scores)#, color=sns.color_palette('Greens')[3])
#plt.ylabel('$\mathcal{T}_F$')
#plt.xticks([1,2,3], [10, 15, 30])
#plt.xlabel('$w$ (seconds)')
plt.tight_layout()
#plt.savefig('../paper/figs/salami-olda-stats.pdf', pad_inches=0.0, bbox_inches='tight')
#!convert ../paper/figs/salami-olda-stats.pdf ../paper/figs/salami-olda-stats.eps

	T_F @ 10	T_P @ 10	T_R @ 10	T_F @ 15	T_P @ 15	T_R @ 15	T_F @ 30	T_P @ 30	T_R @ 30
count	738.000	738.000	738.000	738.000	738.000	738.000	738.000	738.000	738.000
mean	0.363	0.344	0.466	0.391	0.333	0.570	0.367	0.282	0.681
std	0.149	0.194	0.092	0.156	0.180	0.090	0.173	0.176	0.099
min	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
25%	0.282	0.207	0.420	0.300	0.208	0.529	0.244	0.149	0.635
50%	0.381	0.336	0.462	0.405	0.322	0.571	0.357	0.247	0.689
75%	0.462	0.461	0.512	0.495	0.446	0.621	0.491	0.391	0.739
max	0.820	1.000	0.857	0.817	0.986	0.855	0.871	0.935	0.922

	t_measure10	t_measure15	t_measure30	t_precision10	t_precision15	t_precision30	t_recall10	t_recall15	t_recall30
track_id
SALAMI_10	0.337	0.354	0.213	0.242	0.244	0.126	0.558	0.644	0.698
SALAMI_100	0.387	0.503	0.520	0.317	0.438	0.407	0.496	0.592	0.721
SALAMI_1000	0.374	0.377	0.394	0.331	0.306	0.290	0.430	0.491	0.611
SALAMI_1002	0.430	0.418	0.054	0.361	0.317	0.028	0.531	0.614	0.689
SALAMI_1004	0.192	0.162	0.120	0.119	0.093	0.065	0.498	0.640	0.791