/
PimaPrediction.py
169 lines (155 loc) · 6.31 KB
/
PimaPrediction.py
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
import PySimpleGUI as sg
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg
def read_table():
sg.set_options(auto_size_buttons=True)
layout = [[sg.Text('Dataset (a CSV file)', size=(16, 1)),sg.InputText(),
sg.FileBrowse(file_types=(("CSV Files", "*.csv"),("Text Files", "*.txt")))],
[sg.Submit(), sg.Cancel()]]
window1 = sg.Window('Input file', layout)
try:
event, values = window1.read()
window1.close()
except:
window1.close()
return
filename = values[0]
if filename == '':
return
data = []
header_list = []
if filename is not None:
fn = filename.split('/')[-1]
try:
if colnames_checked:
df = pd.read_csv(filename, sep=',', engine='python')
# Uses the first row (which should be column names) as columns names
header_list = list(df.columns)
# Drops the first row in the table (otherwise the header names and the first row will be the same)
data = df[1:].values.tolist()
else:
df = pd.read_csv(filename, sep=',', engine='python', header=None)
# Creates columns names for each column ('column0', 'column1', etc)
header_list = ['column' + str(x) for x in range(len(df.iloc[0]))]
df.columns = header_list
# read everything else into a list of rows
data = df.values.tolist()
# NaN drop?
if dropnan_checked:
df = df.dropna()
data = df.values.tolist()
window1.close()
return (df,data, header_list,fn)
except:
sg.popup_error('Error reading file')
window1.close()
return
def show_table(data, header_list, fn):
layout = [
[sg.Table(values=data,
headings=header_list,
font='Helvetica',
pad=(25,25),
display_row_numbers=False,
auto_size_columns=True,
num_rows=min(25, len(data)))]
]
window = sg.Window(fn, layout, grab_anywhere=False)
event, values = window.read()
window.close()
def show_stats(df):
stats = df.describe().T
header_list = list(stats.columns)
data = stats.values.tolist()
for i,d in enumerate(data):
d.insert(0,list(stats.index)[i])
header_list=['Feature']+header_list
layout = [
[sg.Table(values=data,
headings=header_list,
font='Helvetica',
pad=(10,10),
display_row_numbers=False,
auto_size_columns=True,
num_rows=min(25, len(data)))]
]
window = sg.Window("Statistics", layout, grab_anywhere=False)
event, values = window.read()
window.close()
def sklearn_model(output_var):
"""
Builds and fits a ML model
"""
from sklearn.ensemble import RandomForestClassifier
X = df.drop([output_var], axis=1)
y = df[output_var]
clf = RandomForestClassifier(n_estimators=20,
max_depth=4)
clf.fit(X, y)
#print("Prediction accuracy {}".format(clf.score(X,y)))
return clf, np.round(clf.score(X,y),3)
#=====================================================#
# Define the window's contents i.e. layout
layout = [
[sg.Button('Load data',size=(10,1), enable_events=True, key='-READ-', font='Helvetica 16'),
sg.Checkbox('Has column names?', size=(15,1), key='colnames-check',default=True),
sg.Checkbox('Drop NaN entries?', size=(15,1), key='drop-nan',default=True)],
[sg.Button('Show data',size=(10,1),enable_events=True, key='-SHOW-', font='Helvetica 16',),
sg.Button('Show stats',size=(15,1),enable_events=True, key='-STATS-', font='Helvetica 16',)],
[sg.Text("", size=(50,1),key='-loaded-', pad=(5,5), font='Helvetica 14'),],
[sg.Text("Select output column",size=(18,1), pad=(5,5), font='Helvetica 12'),],
[sg.Listbox(values=(''), key='colnames',size=(30,3),enable_events=True),],
[sg.Text("", size=(50,1),key='-prediction-', pad=(5,5), font='Helvetica 12')],
[sg.ProgressBar(50, orientation='h', size=(100,20), key='progressbar')],
]
# Create the window
window = sg.Window('Pima', layout, size=(600,300))
progress_bar = window['progressbar']
prediction_text = window['-prediction-']
colnames_checked = False
dropnan_checked = False
read_successful = False
# Event loop
while True:
event, values = window.read()
loaded_text = window['-loaded-']
if event in (sg.WIN_CLOSED, 'Exit'):
break
if event == '-READ-':
if values['colnames-check']==True:
colnames_checked=True
if values['drop-nan']==True:
dropnan_checked=True
try:
df,data, header_list,fn = read_table()
read_successful = True
except:
pass
if read_successful:
loaded_text.update("Datset loaded: '{}'".format(fn))
col_vals = [i for i in df.columns]
window.Element('colnames').Update(values=col_vals, )
if event == '-SHOW-':
if read_successful:
show_table(data,header_list,fn)
else:
loaded_text.update("No dataset was loaded")
if event=='-STATS-':
if read_successful:
show_stats(df)
else:
loaded_text.update("No dataset was loaded")
if event=='colnames':
if len(values['colnames'])!=0:
output_var = values['colnames'][0]
if output_var!='Class variable':
sg.Popup("Wrong output column selected!", title='Wrong',font="Helvetica 14")
else:
prediction_text.update("Fitting model...")
for i in range(50):
event, values = window.read(timeout=10)
progress_bar.UpdateBar(i + 1)
_,score = sklearn_model(output_var)
prediction_text.update("Accuracy of Random Forest model is: {}".format(score))