TCSS555B_team17/TCSS555 at master · WebDataScience/TCSS555B_team17 · 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
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
import pandas as pd
import csv
from sklearn.linear_model import LinearRegression
from sklearn.cross_validation import train_test_split
import argparse
from xml.etree import ElementTree
import sys
from sklearn.ensemble import RandomForestClassifier
import warnings
warnings.filterwarnings("ignore", category=DeprecationWarning)


def linearRegr(train_data):
    feature_cols = ['WC', 'WPS', 'Sixltr', 'Dic', 'Numerals', 'funct', 'pronoun', 'ppron', 'i', 'we', 'you', 'shehe', 'they', 'ipron', 'article', 'verb', 'auxverb', 'past', 'present', 'future', 'adverb', 'preps', 'conj', 'negate', 'quant', 'number', 'swear', 'social', 'family', 'friend', 'humans', 'affect', 'posemo', 'negemo', 'anx', 'anger', 'sad', 'cogmech', 'insight', 'cause', 'discrep', 'tentat', 'certain', 'inhib', 'incl', 'excl', 'percept', 'see', 'hear', 'feel', 'bio', 'body', 'health', 'sexual', 'ingest', 'relativ', 'motion', 'space', 'time', 'work', 'achieve', 'leisure', 'home', 'money', 'relig', 'death', 'assent', 'nonfl', 'filler', 'Period', 'Comma', 'Colon', 'SemiC', 'QMark', 'Exclam', 'Dash', 'Quote', 'Apostro', 'Parenth', 'OtherP', 'AllPct']
    X = train_data[feature_cols]

    y1 = train_data.ope
    lm1 = LinearRegression()
    lm1.fit(X, y1)
    zip(feature_cols, lm1.coef_)
    y2 = train_data.con
    lm2 = LinearRegression()
    lm2.fit(X, y2)
    zip(feature_cols, lm2.coef_)
    y3 = train_data.ext
    lm3 = LinearRegression()
    lm3.fit(X, y3)
    zip(feature_cols, lm3.coef_)
    y4 = train_data.agr
    lm4 = LinearRegression()
    lm4.fit(X, y4)
    zip(feature_cols, lm4.coef_)
    y5 = train_data.neu
    lm5 = LinearRegression()
    lm5.fit(X, y5)
    zip(feature_cols, lm5.coef_)
    lr = [lm1,lm2,lm3,lm4,lm5]
    return lr


def train_agen_traits(merged):

    ageCat = 0
    for i in range(len(merged)):
        a=merged.iloc[i].age
        if a <=24:
            ageCat = 1
        elif a >= 25 and a <=34:
            ageCat = 2
        elif a >= 35 and a <=49:
            ageCat = 3
        elif a >= 50 :
            ageCat = 4
        else: print ("invalid age group")
        merged.loc[i,'age'] = ageCat

    feature_cols = ['WC', 'WPS', 'Sixltr', 'Dic', 'Numerals', 'funct', 'pronoun', 'ppron', 'i', 'we', 'you', 'shehe', 'they', 'ipron', 'article', 'verb', 'auxverb', 'past', 'present', 'future', 'adverb', 'preps', 'conj', 'negate', 'quant', 'number', 'swear', 'social', 'family', 'friend', 'humans', 'affect', 'posemo', 'negemo', 'anx', 'anger', 'sad', 'cogmech', 'insight', 'cause', 'discrep', 'tentat', 'certain', 'inhib', 'incl', 'excl', 'percept', 'see', 'hear', 'feel', 'bio', 'body', 'health', 'sexual', 'ingest', 'relativ', 'motion', 'space', 'time', 'work', 'achieve', 'leisure', 'home', 'money', 'relig', 'death', 'assent', 'nonfl', 'filler', 'Period', 'Comma', 'Colon', 'SemiC', 'QMark', 'Exclam', 'Dash', 'Quote', 'Apostro', 'Parenth', 'OtherP', 'AllPct']
    target_age = merged.age
    train = merged[feature_cols]
    target_gen = merged.gender

    rf_age = RandomForestClassifier(n_estimators=100)
    rf_gen = RandomForestClassifier(n_estimators=100)
    rf_age.fit(train, target_age)
    rf_gen.fit(train, target_gen)

    l_obj = linearRegr(merged)

    return rf_gen,rf_age,l_obj


def parse_args():
    parser = argparse.ArgumentParser(description="""Script takes full input path to
                         test directory, output directory and training directory""")

    parser.add_argument('-i',
                        "--test_dir",
                        type=str,
                        required=True,
                        help='Full path to input test directory containing profile and text dir')

    parser.add_argument('-o', "--output_dir",
                        type=str,
                        required=True,
                        help='The path to output directory')


    args = parser.parse_args()
    return args

#PROGRAM START
if len(sys.argv) < 2:
    print "Invalid arguments!"
    exit()

args = parse_args()

input_dir = args.test_dir
output_dir = args.output_dir

print(input_dir,output_dir)

#Training data
# read profile data into dataframe
data1 = pd.read_csv("/data/training/profile/profile.csv", index_col=0)
# read LIWC data into a DataFrame
data2 = pd.read_csv('/data/training/LIWC.csv', index_col=1)
merged = pd.merge(left=data1,right=data2, how='left',left_on='userid',right_on= 'userId')

#test data preparation
# read profile data into dataframe
datatest1 = pd.read_csv(input_dir+'/profile/profile.csv', index_col=0)
# read LIWC data into a DataFrame
datatest2 = pd.read_csv(input_dir+'/LIWC.csv', index_col=1)

merged_test = pd.merge(left=datatest1,right=datatest2, how='left',left_on='userid',right_on= 'userId')

#Train the classifiers
rf_gen,rf_age,l_obj = train_agen_traits(merged)
#Test
feature_cols = ['WC', 'WPS', 'Sixltr', 'Dic', 'Numerals', 'funct', 'pronoun', 'ppron', 'i', 'we', 'you', 'shehe', 'they', 'ipron', 'article', 'verb', 'auxverb', 'past', 'present', 'future', 'adverb', 'preps', 'conj', 'negate', 'quant', 'number', 'swear', 'social', 'family', 'friend', 'humans', 'affect', 'posemo', 'negemo', 'anx', 'anger', 'sad', 'cogmech', 'insight', 'cause', 'discrep', 'tentat', 'certain', 'inhib', 'incl', 'excl', 'percept', 'see', 'hear', 'feel', 'bio', 'body', 'health', 'sexual', 'ingest', 'relativ', 'motion', 'space', 'time', 'work', 'achieve', 'leisure', 'home', 'money', 'relig', 'death', 'assent', 'nonfl', 'filler', 'Period', 'Comma', 'Colon', 'SemiC', 'QMark', 'Exclam', 'Dash', 'Quote', 'Apostro', 'Parenth', 'OtherP', 'AllPct']

xtest = merged_test[feature_cols]
xtest = xtest.copy()

for i in range(len(merged_test)):
    x = rf_age.predict(xtest.iloc[i])
    y = rf_gen.predict(xtest.iloc[i])
    if y==0:
        gender_cat = "male"
    else:
        gender_cat = "female"

    ageGroup = " "
    if x <=24:
        ageGroup = "xx to 24"
    elif x >= 25 and x <=34:
        ageGroup = "25 to 34"
    elif x >= 35 and x <=49:
        ageGroup = "35 to 49"
    elif a >= 50 :
        ageGroup = "50 to xx"

    xo=l_obj[0].predict(xtest.iloc[i])
    xc=l_obj[1].predict(xtest.iloc[i])
    xe=l_obj[2].predict(xtest.iloc[i])
    xa=l_obj[3].predict(xtest.iloc[i])
    xn=l_obj[4].predict(xtest.iloc[i])

    user = merged_test.iloc[i].userid
    output_file = output_dir+user+".xml"
    with open(output_file, "w") as out_f:
        attrs = {"userId": user,
                 "gender" : gender_cat,
                 "age_group" : ageGroup,
                 "extrovert" : str(round(xe,1)),
                 "neurotic" : str(round(xn,1)),
                 "agreeable" : str(round(xa,1)),
                 "conscientious" : str(round(xc,1)),
                 "open" : str(round(xo,1))
                }
        tree = ElementTree.Element('', attrs)
        out_f.write(ElementTree.tostring(tree))