-
Notifications
You must be signed in to change notification settings - Fork 0
/
yolo.py
49 lines (45 loc) · 1.77 KB
/
yolo.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
# -*- coding: utf-8 -*-
"""
Created on Fri Jul 19 16:42:20 2013
@author: Bit Shift Boys
"""
import pandas as pd
# import rows 1-4501 and add column titles
train = pd.read_csv('./input00.txt', sep="\s", skiprows=1,nrows=4500, names=range(25))
# import rows 4502-6002 and add column titles
test = pd.read_csv('./input00.txt', sep="\s", skiprows=4502, names=range(24))
# for columns slice out col id and :
# training data - first 9 features
for col in range(2,11):
train[col] = train[col].map(lambda val: float(str(val)[2:]))
# training data - first 9 features
for col in range(11,25):
train[col] = train[col].map(lambda val: float(str(val)[3:]))
# test data - first 9 features
for col in range(1,10):
test[col] = test[col].map(lambda val: float(str(val)[2:]))
# test data - first 9 features
for col in range(10,24):
test[col] = test[col].map(lambda val: float(str(val)[3:]))
# scale data
# turn -1 to 0 and leave 1 as 1 - THIS WILL TURN ALL NON -1 VALS TO 1!!! check this
train[1] = train[1].map(lambda val: 0 if val == -1 else 1)
test[1] = test[1].map(lambda val: 0 if val == -1 else 1)
# find max and divide each number by max
for col in range(2,25):
max = train[col].max()
if max != 0: train[col] = train[col].map(lambda val: val / max)
for col in range(2,24):
max = test[col].max()
if max != 0: test[col] = test[col].map(lambda val: val / max)
# move target column to end (only in train)
train[25] = train[1]
for col in range(2,26):
train[col-1] = train[col]
del train[25]
# output csv files
train.to_csv('./train.csv', header=False, index=False)
test.to_csv('./test.csv', header=False, index=False)
# delete features and outliers based on visualizations
train_plucked = train.drop([2182,3460,4057,4080])
train_plucked.to_csv('./train-plucked.csv', header=False, index=False)