TAMUDatathon_eatsy/modeltrain.py at master · Sanghyeon16/TAMUDatathon_eatsy · 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
164
165
166
167
168
169
#!/usr/bin/env python
# coding: utf-8

# In[190]:


import json
import ast
import pandas as pd
from collections import defaultdict
import time
import numpy as np
import scipy.stats
from sklearn.linear_model import LinearRegression
from sklearn.linear_model.logistic import LogisticRegression
from sklearn.preprocessing import PolynomialFeatures
from sklearn.linear_model.logistic import _logistic_regression_path
import matplotlib.pyplot as plt


# In[550]:


with open('buisnessoutput.txt') as json_file:
    bo = json.load(json_file)
with open('users.txt') as json_file:
    users = json.load(json_file)


# In[551]:


with open('allattrs.txt') as json_file:
    allattrs = json.load(json_file)
with open('allcats.txt') as json_file:
    allcats = json.load(json_file)
allca = allcats + allattrs
len(allca)
with open('allca.txt', 'w') as outfile:
    json.dump(list(sorted(allca)),outfile)


# In[552]:


def vecsperuser(usr):
    vecs={}
#     print(usr)
    j = 0
    inp2 = list(users[usr].keys())
#     print(inp2)
    for bsn in inp2:
#         print ('              ',end = '\r')
#         print (round(100*i / len(inp),2),round((100*j / len(inp2)),2),end='\r')
        print (round(100*j / len(inp2),2),end='\r')
#         print(bsn)
        savesum = bo[bsn]
        newsum = {}
#         print('got here 1')
        if not len(savesum) == 0:
#             print('got here 2')
            sumset = set(sorted(savesum.keys()))
            catset = set([x for x in list(sumset) if x[:10] == 'categories'])
            resset = sumset - catset
#             print(len(savesum),'savesum',savesum)
#             print(len(catset),'catset',catset)
#             print(len(resset),'resset',resset)
            #diff = set(allattrs) - resset
            diff = set(allca) - sumset

#             print(len(diff),'diff')
            for differing in diff:
                newsum[differing] = False
            for res in sumset:
                newsum[res] = savesum[res]
    #         sortedkeys = sorted(list(savesum.keys()))
            if len(newsum) != 831:
                print(bsn)
                print(len(newsum),'res')
                print(sorted(list(newsum.keys())),'\n')
                print(savesum.keys(),'\n')
#             print('\n')
#             print(np.asarray(sorted(newsum.items()))[:,1])
            temp = [1 if x == "True" else 0 for x in np.asarray(sorted(newsum.items()))[:,1]]
#             print(temp)
            vecs[bsn] = temp
#         print('got here 3')
        j += 1
    return vecs


# In[561]:


def findgoodinps_helper(inp):
    vinp = vecsperuser(inp)
    vecs = np.asarray(list(vinp.values()))
    label = []
    for bsn in vinp.keys():
        users[inp]
        label.append(users[inp][bsn])

    if len(label) > 10 and not min(label) == max(label):
#         print('\r')
        return inp
#     c = LogisticRegression(random_state=0, solver='lbfgs').fit(vecs, label).coef_


# In[566]:


def findgoodinps(limit = 100):
    goodinp = []
    for inp in list(users.keys())[0:limit]:
        goodinp.append(findgoodinps_helper(inp))
    goodinp = [x for x in goodinp if not x == None]
    return goodinp


# In[531]:

def make_pretty(attr):
    pretty_string = ""
    for i in range(len(attr)):
        char = attr[i]

        if char == '_':
            pretty_string += ' '
            continue

        if char.isupper() and not attr[i - 1].isupper() and not attr[i + 1].isupper() and i != 0:
            pretty_string += ' '

        pretty_string += char

    return pretty_string


def makemodel(inp):
    tojson = {}
    vinp = vecsperuser(inp)
    vecs = np.asarray(list(vinp.values()))
    label = []
    for bsn in vinp.keys():
        users[inp]
        label.append(users[inp][bsn])
    c = LogisticRegression(random_state=0, solver='lbfgs').fit(vecs, label).coef_[0]
    arr = np.asarray([round(i,2) for i in c])
    arr = (arr-min(arr))/(max(arr)-min(arr))
    marr = (arr+1)*100000
#     print(marr)
#     print(max(marr),min(marr))
    top_idx = np.argsort(c)[-10:]
    top_values = [c[i] for i in top_idx]
    for idx in top_idx:
        tojson[list(sorted(allca))[idx]] =int(round( marr[idx],0))
    arr2 = np.asarray(list(tojson.values()))
    arr2 = (arr2-min(arr2))/(max(arr2)-min(arr2)) * 10
    print(arr2)
    arr2 = [int(round( i,0)) for i in arr2]

    oldkey = lambda i: list(tojson.keys())[i]
    return c, {make_pretty(oldkey(i)):arr2[9 - i] for i in range(3)}


# In[532]:


# c = makemodel(goodinp[2])