In your face:
Biased judgement of fear-anger expressions in violence offenders.

4. Basic Plotting and Statistics for morphed faces

import libraries

In [1]:
import numpy as np
import pandas as pd

import os
import fnmatch

import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline

from myBasics import *
In [2]:
bigDf = pd.read_csv('../outputs/genderTable.csv',
                    index_col=[0,1,2])
In [3]:
bigDf.head()
Out[3]:
00 01 02 03 04 05 06 07 08 09 10 group
g p fgender
A A001 F 0.00 0.10 0.10 0.2 0.15 0.30 0.85 0.90 0.90 1.00 1.00 2
M 0.00 0.10 0.20 0.2 0.25 0.50 0.70 0.85 0.95 1.00 1.00 2
A002 F 0.20 0.10 0.05 0.3 0.25 0.50 0.75 0.90 0.75 1.00 1.00 2
M 0.05 0.25 0.35 0.2 0.35 0.55 0.60 0.90 0.85 1.00 1.00 2
A003 F 0.00 0.05 0.15 0.1 0.30 0.40 0.55 0.65 0.65 0.65 0.85 2

One plot for each face gender

In [4]:
def makeGenderSpaghetti(bigDf,cond,count):
    
    ax = plt.subplot(1,2,count)
    
    # to loop through all cases but have only one legend without redundancies,
    # here we keep track of whether a condition is already labeled
    legTrack = []
    
    # looping  through the 3 groups
    for group in bigDf.index.levels[0]:
        # looping through the participants of each group
        for p in bigDf.ix[group].index:
            if p[0][0] in group:
                
                # color and legend setting for that group
                thisCol = myPal[labelCoding[p[0][0]]]
                thisLeg = myGLabels[p[0][0]]
                
                # get the data from the specified row
                thisList = list(bigDf.ix[group].ix[p[0]].ix[cond])[:-1] # last row is group membership; get rid of that

                # plotting with legend if this is the first instance,
                # otherwise plot without a legend
                if thisLeg not in legTrack:
                    ax.plot(thisList,
                            c=thisCol,
                            alpha=0.5,
                            linewidth=4,
                            label=thisLeg)
                else:
                    ax.plot(thisList,
                            c=thisCol,
                            linewidth=4,
                            alpha=0.5)            
                # keep track which conditions already have a legend
                legTrack.append(thisLeg)
        
    # plot formatting

    ax.set_xlabel('Morphing Grade (Fear --> Anger)')
    ax.set_ylabel('% Anger Responses')

    plt.xticks(np.arange(0,10.1,2), [str(a)+'%' for a in  np.arange(0,101,20)])
    plt.yticks(np.arange(0,1.01,0.2), [str(a)+'%' for a in  np.arange(0,101,20)])

    #plt.ylim(.0,1.); plt.xlim(-0.1,10.2)
    if cond == 'M':
        ax.set_title('Male Faces')
    elif cond == 'F':
        ax.set_title('Female Faces')
        ax.set_ylabel('')
        ax.set_yticks([])
        # one legend for both plots
        plt.legend(loc='best',bbox_to_anchor=[1.1, 1])
In [5]:
plt.figure(figsize=(12,4))
for i,fgender in enumerate(['M','F']):
    makeGenderSpaghetti(bigDf,fgender,i+1)
    sns.despine()
#plt.savefig('../figures/rawSpaghettiAll.png',dpi=300,bbox_inches="tight")
plt.show()

Same thing, but as interactive plot

In [6]:
def interactiveSpaghetti(i):
    
    mmDf = bigDf[1::2]
    mDf = mmDf[mmDf.columns[:-1]]
    mY = list( mmDf['group'] )
    
    ffDf = bigDf[0::2]
    fDf = ffDf[ffDf.columns[:-1]]
    fY = list( ffDf['group'])
  
    
    fig = plt.figure(figsize=(12,4))
    
    ### male
    ax = plt.subplot(1,2,1)
    
    selectDf = mDf.ix[i]
    otherDf = np.array( pd.concat([mDf[:i],mDf[i+1:] ] ) )
        
    for entry in range(otherDf.shape[0]):
        try: ax.plot(otherDf[entry],alpha=0.5,color=myPal[mY[entry]])
        except: print "!"
            
    ax.plot(selectDf,linewidth=10,color=myPal[mY[i]])
    
    ### female
    ax = plt.subplot(1,2,2)
    
    selectDf = fDf.ix[i]
    otherDf = np.array( pd.concat([fDf[:i],fDf[i+1:] ] ) )
        
    for entry in range(otherDf.shape[0]):
        try: ax.plot(otherDf[entry],alpha=0.5,color=myPal[fY[entry]])
        except: print "!"
            
    ax.plot(selectDf,linewidth=10,color=myPal[fY[i]])
    sns.despine()
    
    return fig
In [7]:
# Interactive plots for static html notebooks; using ipywidgets by Jake Vanderplas  
# https://github.com/jakevdp/ipywidgets-static

from ipywidgets_static import StaticInteract, RangeWidget, RadioWidget
In [8]:
StaticInteract(interactiveSpaghetti,
               i=RangeWidget(0, 60, 1)
              )

/opt/anaconda2/lib/python2.7/site-packages/matplotlib/pyplot.py:516: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  max_open_warning, RuntimeWarning)
Out[8]:
i:

get rid of outliers

Here, some participants are removed whose response functions indicate guessing or non-compliance.
We also visualise the data of the excluded participants here, to make clear why they had to be excluded.

In [9]:
eliminated = ['G011','G013','G016','G025','A004']
In [10]:
fig = plt.figure(figsize=(16,8))
for e,elim in enumerate(eliminated):
    ax = plt.subplot(2,len(eliminated),e+1)
    ax.plot( bigDf.ix[elim[0]].ix[elim].ix['M'][:-1] )
    ax.set_title(elim+' M')
    ax.set_ylim(0,1)
for e,elim in enumerate(eliminated):
    ax = plt.subplot(2,len(eliminated),len(eliminated)+e+1)
    ax.plot( bigDf.ix[elim[0]].ix[elim].ix['F'][:-1] )
    ax.set_title(elim+' F')
    ax.set_ylim(0,1)
    sns.despine()
plt.tight_layout()
plt.show()

plot cleaned up data

In [11]:
for elim in eliminated:
    bigDf.drop([elim],level=1, inplace=True)
In [12]:
plt.figure(figsize=(12,4))
for i,fgender in enumerate(['M','F']):
    makeGenderSpaghetti(bigDf,fgender,i+1)
    sns.despine()
#plt.savefig('../figures/rawSpaghetti.png',dpi=300,bbox_inches="tight")
plt.show()

Changing the structure of the DataFrame to easily average over group

In [13]:
genderDf = bigDf.unstack(0).unstack(0).stack(1).stack(1)
In [14]:
genderDf.head()
Out[14]:
00 01 02 03 04 05 06 07 08 09 10 group
fgender g p
F A A001 0.00 0.10 0.10 0.20 0.15 0.30 0.85 0.90 0.90 1.00 1.00 2.0
A002 0.20 0.10 0.05 0.30 0.25 0.50 0.75 0.90 0.75 1.00 1.00 2.0
A003 0.00 0.05 0.15 0.10 0.30 0.40 0.55 0.65 0.65 0.65 0.85 2.0
A005 0.35 0.20 0.15 0.55 0.40 0.70 0.90 1.00 1.00 1.00 1.00 2.0
A006 0.00 0.20 0.05 0.25 0.20 0.35 0.65 0.55 0.85 0.85 0.95 2.0
In [15]:
# get rid of group column
genderDf = genderDf[genderDf.columns[:-1]]
In [16]:
genderDf.head()
Out[16]:
00 01 02 03 04 05 06 07 08 09 10
fgender g p
F A A001 0.00 0.10 0.10 0.20 0.15 0.30 0.85 0.90 0.90 1.00 1.00
A002 0.20 0.10 0.05 0.30 0.25 0.50 0.75 0.90 0.75 1.00 1.00
A003 0.00 0.05 0.15 0.10 0.30 0.40 0.55 0.65 0.65 0.65 0.85
A005 0.35 0.20 0.15 0.55 0.40 0.70 0.90 1.00 1.00 1.00 1.00
A006 0.00 0.20 0.05 0.25 0.20 0.35 0.65 0.55 0.85 0.85 0.95
In [17]:
def makePlot(genderDf,groups):
    plt.figure(figsize=(12,4))

    for count,gender in enumerate( genderDf.index.levels[0][::-1] ):

        ax = plt.subplot(1,2,count+1)

        for j,group in enumerate( groups ):

            thisMean = genderDf.ix[gender].ix[group].mean()
            thisStd = genderDf.ix[gender].ix[group].std(ddof=1)

            thisCol = myPal[labelCoding[group]]
            thisLeg = myGLabels[group]
            
            plt.errorbar( np.linspace(0,10,len(thisMean))+j/15., # apply small shift so CI are better visible
                          thisMean,
                          yerr=thisStd/np.sqrt(len(genderDf.ix[gender].ix[group]) )*1.96,

                          label=thisLeg,
                          color=thisCol,
                          linewidth=3,
                          alpha=0.7
                        )

        # plot formatting

        ax.set_xlabel('Morphing Grade (Fear --> Anger)')
        ax.set_ylabel('% Anger Responses')

        plt.xticks(np.arange(0,10.1,2), [str(a)+'%' for a in  np.arange(0,101,20)])
        plt.yticks(np.arange(0,1.01,0.2), [str(a)+'%' for a in  np.arange(0,101,20)])

        plt.ylim(-0.02,1.02); plt.xlim(-0.1,10.2)
        if gender == 'M':
            ax.set_title('Male Faces')
        elif gender == 'F':
            ax.set_title('Female Faces')
            ax.set_ylabel('')
            ax.set_yticks([])
            # one legend for both plots
            plt.legend(loc='best',bbox_to_anchor=[1, 1])

    #plt.show()
In [18]:
makePlot(genderDf,['G','A'])
sns.despine()
#plt.savefig('../figures/meanGenderPlotViolenceGeneral.png',dpi=300,bbox_inches="tight")
plt.show()
In [19]:
makePlot(genderDf,['G','K'])
sns.despine()
#plt.savefig('../figures/meanGenderPlotViolenceMolesters.png',dpi=300,bbox_inches="tight")
plt.show()

Restructure for seaborn plots

In [20]:
genderStacked = genderDf.unstack(0).stack(0).unstack(2)
In [21]:
groupCol = []
for entry in genderStacked.index.levels[1]:
    try:
        genderStacked.ix[entry[0]].ix[entry]
        groupCol.append(labelCoding[entry[0]] )
    except:
        print entry # previously eliminated cases are excluded
genderStacked['group'] = groupCol
In [22]:
genderStacked.head()
Out[22]:
fgender F ... M group
00 01 02 03 04 05 06 07 08 09 ... 02 03 04 05 06 07 08 09 10
g p
A A001 0.00 0.10 0.10 0.20 0.15 0.30 0.85 0.90 0.90 1.00 ... 0.20 0.20 0.25 0.50 0.70 0.85 0.95 1.00 1.00 2
A002 0.20 0.10 0.05 0.30 0.25 0.50 0.75 0.90 0.75 1.00 ... 0.35 0.20 0.35 0.55 0.60 0.90 0.85 1.00 1.00 2
A003 0.00 0.05 0.15 0.10 0.30 0.40 0.55 0.65 0.65 0.65 ... 0.05 0.05 0.30 0.25 0.45 0.50 0.80 0.95 0.95 2
A005 0.35 0.20 0.15 0.55 0.40 0.70 0.90 1.00 1.00 1.00 ... 0.30 0.45 0.40 0.65 0.90 1.00 1.00 1.00 1.00 2
A006 0.00 0.20 0.05 0.25 0.20 0.35 0.65 0.55 0.85 0.85 ... 0.30 0.30 0.45 0.50 0.85 0.85 0.90 0.90 1.00 2

5 rows × 23 columns

In [23]:
fig = plt.figure(figsize=(12,6))

for i,fGender in enumerate(['M','F']):
    ax = fig.add_subplot('12'+str(i+1))

    sns.boxplot(x='group',y=(fGender,'05'),data=genderStacked,
                    width=0.4,linewidth=1,color='white',whis=True,notch=True,fliersize=0,ax=ax)
    sns.stripplot(x='group',y=(fGender,'05'),data=genderStacked,
                      jitter=True, edgecolor='white',palette=myPal,size=9,linewidth=1,ax=ax)

    if fGender == 'M':
        ax.set_title('Male Faces')
    else :
        ax.set_title('Female Faces')        
        
    ax.set_ylim(0,1)
    ax.set_xticklabels(['Violence\nOffenders','Child\nMolesters','General\nPopulation'],fontsize=15)
    ax.set_xlabel('')
    if i==0:
        ax.set_ylabel('Anger Responses')
    else:
        ax.set_ylabel('')
    sns.despine()
    plt.yticks(np.arange(0,1.01,0.1), [str(a)+'%' for a in  np.arange(0,101,10)])
#plt.suptitle('Threshold of fitted function',fontsize=20, position=(0.51,1.1))
plt.tight_layout()
#plt.savefig('../figures/fifty_fifty.png',dpi=300)
plt.show()

inferential statistics

In [24]:
from scipy import stats
In [25]:
def makeMannUTests(df):
    bigDf = pd.DataFrame()
    d = {}
    for comp in [('G','K'),('G','A'),('K','A')]:
        for variable in df.columns:
            group1 = df.ix[comp[0]][variable]
            group2 = df.ix[comp[1]][variable]
            U,p = stats.mannwhitneyu(group1,group2)
            if p<0.05:
                thisSig = '*'
            else:
                thisSig = 'n.s.'
            d[variable] = {'U':round(U,2),'p':round(p,3),'sig':thisSig}
        thisDf = pd.DataFrame(d)
        thisDf = thisDf.reindex_axis(['U','p','sig'], axis=0)
        thisDf.index = [ [comp[0]+' > '+comp[1]]*len(thisDf.index),thisDf.index ]

        bigDf = pd.concat([bigDf,thisDf])
    return bigDf.T
In [26]:
makeMannUTests(genderDf.unstack(0))[1::2]
Out[26]:
G > K G > A K > A
U p sig U p sig U p sig
00 M 207.5 0.473 n.s. 293.5 0.026 * 144.5 0.177 n.s.
01 M 211.5 0.408 n.s. 280.5 0.06 n.s. 123 0.659 n.s.
02 M 207.5 0.475 n.s. 283.5 0.05 n.s. 131.5 0.424 n.s.
03 M 195.5 0.709 n.s. 312 0.007 * 156.5 0.064 n.s.
04 M 218 0.309 n.s. 325.5 0.002 * 155 0.073 n.s.
05 M 286.5 0.003 * 352.5 0 * 133.5 0.377 n.s.
06 M 244 0.077 n.s. 302 0.015 * 122 0.691 n.s.
07 M 255.5 0.036 * 250.5 0.271 n.s. 96 0.516 n.s.
08 M 196.5 0.686 n.s. 286.5 0.039 * 143 0.2 n.s.
09 M 201.5 0.572 n.s. 252 0.241 n.s. 128 0.507 n.s.
10 M 227 0.167 n.s. 279 0.049 * 118 0.812 n.s.
In [27]:
makeMannUTests(genderDf.unstack(0))[::2]
Out[27]:
G > K G > A K > A
U p sig U p sig U p sig
00 F 218 0.306 n.s. 290 0.03 * 140 0.23 n.s.
01 F 228.5 0.185 n.s. 264 0.142 n.s. 117 0.847 n.s.
02 F 206 0.5 n.s. 252 0.254 n.s. 124 0.626 n.s.
03 F 201 0.597 n.s. 238 0.438 n.s. 110.5 0.967 n.s.
04 F 209 0.448 n.s. 273 0.092 n.s. 132 0.4 n.s.
05 F 222.5 0.254 n.s. 298 0.02 * 136 0.326 n.s.
06 F 238 0.114 n.s. 276.5 0.077 n.s. 126 0.572 n.s.
07 F 234.5 0.138 n.s. 284 0.049 * 130.5 0.45 n.s.
08 F 240.5 0.095 n.s. 313 0.006 * 146 0.158 n.s.
09 F 178 0.918 n.s. 254.5 0.22 n.s. 141.5 0.216 n.s.
10 F 202 0.56 n.s. 240 0.391 n.s. 118.5 0.796 n.s.

save with new header, for ANOVA in JASP

In [28]:
jaspGender = genderDf.unstack(0)
jaspGender.columns = jaspGender.columns.swaplevel(0, 1)
jaspGender.sortlevel(0, axis=1, inplace=True)
jaspGender.columns = ['fm'[jaspGender.columns.labels[0][x]]+'_'+str(jaspGender.columns.labels[1][x]) for x in range(jaspGender.shape[1]) ]
In [29]:
jaspGender.head()
Out[29]:
f_0 f_1 f_2 f_3 f_4 f_5 f_6 f_7 f_8 f_9 ... m_1 m_2 m_3 m_4 m_5 m_6 m_7 m_8 m_9 m_10
g p
A A001 0.00 0.10 0.10 0.20 0.15 0.30 0.85 0.90 0.90 1.00 ... 0.10 0.20 0.20 0.25 0.50 0.70 0.85 0.95 1.00 1.00
A002 0.20 0.10 0.05 0.30 0.25 0.50 0.75 0.90 0.75 1.00 ... 0.25 0.35 0.20 0.35 0.55 0.60 0.90 0.85 1.00 1.00
A003 0.00 0.05 0.15 0.10 0.30 0.40 0.55 0.65 0.65 0.65 ... 0.00 0.05 0.05 0.30 0.25 0.45 0.50 0.80 0.95 0.95
A005 0.35 0.20 0.15 0.55 0.40 0.70 0.90 1.00 1.00 1.00 ... 0.20 0.30 0.45 0.40 0.65 0.90 1.00 1.00 1.00 1.00
A006 0.00 0.20 0.05 0.25 0.20 0.35 0.65 0.55 0.85 0.85 ... 0.40 0.30 0.30 0.45 0.50 0.85 0.85 0.90 0.90 1.00

5 rows × 22 columns

In [30]:
jaspGender.to_csv('../outputs/genderMorphsJASP.csv')

Correlation of 50/50 with AFAS

Get the AFAS data

In [31]:
afasDf = pd.read_csv('../outputs/meanAFAS.csv',index_col=[0,1])

Get rid of the participants that were excluded from the morphing task

In [32]:
for elim in eliminated:
    afasDf.drop([elim],level=1, inplace=True)
In [33]:
afasDf.head()
Out[33]:
Facilitative Appetitive Overall group
id
A A001 1.000000 0.266667 0.633333 2
A002 0.266667 0.066667 0.166667 2
A003 0.600000 0.200000 0.400000 2
A005 1.400000 1.200000 1.300000 2
A006 0.466667 0.266667 0.366667 2

Merge together:

In [34]:
assert ( afasDf.index == genderStacked.index ).all()
In [35]:
mergeDf = pd.DataFrame([ afasDf['Overall'], genderStacked['M']['05'], afasDf['group'] ]).T
In [36]:
sns.jointplot("Overall", "05", data=mergeDf, kind="reg",
              xlim=(0, 4), ylim=(0, 1),stat_func=None,
              scatter_kws={"s": 70,"edgecolor":"white","linewidth":1}
             )

#plt.savefig('../figures/afasFiftyCorrelationAll.png',dpi=300)


sns.lmplot(x="Overall", y="05", data=mergeDf,hue="group",ci=None,truncate=True,palette=myPal,
           scatter_kws={"s": 70,"edgecolor":"white","linewidth":1}
          )
plt.ylim(0,1);plt.xlim(0,4)
#plt.savefig('../figures/afasFiftyCorrelationGroups.png',dpi=300)
Out[36]:
(0, 4)
In [37]:
for scale in afasDf.columns[:-1]:
    print "\n",scale
    r, p = stats.spearmanr(afasDf[scale], genderStacked['M']['05'])
    print "Spearman r:",r,"p:",p
    r, p = stats.pearsonr(afasDf[scale], genderStacked['M']['05'])
    print "Pearson r:",r,"p:",p

Facilitative
Spearman r: 0.348177427724 p: 0.00854869903188
Pearson r: 0.423191595328 p: 0.00115564812856

Appetitive
Spearman r: 0.360013818649 p: 0.00642285626403
Pearson r: 0.423972371242 p: 0.00112899301368

Overall
Spearman r: 0.365846912775 p: 0.00555700546699
Pearson r: 0.435032544467 p: 0.000806331193605
In [ ]: