Day 12_RandomForeset_MushroomDataset

# -*- coding: utf-8 -*-
"""Day 14_RandomForest_mushroom.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1jCkmL1FflmAn4pEapKnpZjR5obBqB_MI
"""
 
from google.colab import drive
drive.mount('/gdrive')
 
PATH = "/gdrive/My Drive/Colab Notebooks/resources/"
 
# %matplotlib inline
 
import numpy as np
import matplotlib.pyplot as plt
import matplotlib as mpl
import pandas as pd
import seaborn as sns
import time
 
 
from scipy.stats import norm, multivariate_normal
# installing packages for interactive graphs
import ipywidgets as widgets
from IPython.display import display
from ipywidgets import interact, interactive, fixed, interact_manual, IntSlider
 
##############################################
################ Helper Function##############
##############################################
 
def my_df_dropNas(df, columns):
    for col in columns:
        df =  df[ df[col].notna() ]
    return df
        
def my_checkNas( x ):
    y = x.apply(lambda x : (
    sum(x.isna())
))
    return y
 
@interact_manual( x =IntSlider(0,0,12) )
def test_model( x ):
    print(x)
 
## 웹에서 다운로드
import urllib.request as req
url = "http://archive.ics.uci.edu/ml/machine-learning-databases/mushroom/"+"/agaricus-lepiota.data"
req.urlretrieve(url,"mushdata.csv")
 
df = pd.read_csv("mushdata.csv",header = None)
df.shape
 
"""#각 카테고리를 숫자로 변환"""
 
## 아스키 코드를 기반으로 숫자 <-> 문자 변환
print(ord('a'))
print(chr(97))
 
label = []
data  = []
 
for row_index, row in df.iterrows():
    label.append(row.iloc[0])
    row_data = []
    for v in row.iloc[1:]:
        row_data.append(ord(v))
    
    data.append(row_data)
    
len(label)
 
from sklearn.model_selection import train_test_split
 
## Split test and train data
data_train, data_test, label_train, label_test = train_test_split(data, label)
 
len(data_train) - len(data_test)
 
from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
 
## 데이터 학습
clf = RandomForestClassifier()
 
clf.fit(data_train, label_train)
 
predict = clf.predict(data_test)
 
## 정확도 측정
ac_score = metrics.accuracy_score(label_test, predict)
 
## 여러 측정 방법 적용
rep = metrics.classification_report(label_test, predict)
 
ac_score
 
print(rep)
 
"""데이터 수치화에서 중요한것.

빨강 = 1
파랑 = 2
초록 = 3
흰색 = 4

숫자 사이에 특별한 관련성은 없다.

해당 숫자가 카테고리화를 위한 변수인지, 혹은 수치형 변수인지 잘 생각해야 한다.


빨강 => 1000
파랑 => 0100  카테고리화를 위해서 원핫인코딩을 해야한다.
"""
 
label     = []
data      = []
attr_list = []
 
 
for row_index , row in df.iterrows():
    label.append( row.ix[0] )
    exdata = []
    
    ## col = 열번호 , v = 각 문자
    for col, v in enumerate(row.ix[1:]):
        if row_index == 0:
            
            ## attr = "dic" : {}, "cnt" : 0  ## 딕셔너리 초기화
            
            attr = { "dic" : {}, "cnt" : 0 }
            attr_list.append(attr)
            
        else :
            attr = attr_list[col]
        
        ## 변수의 특징을 표현하는 변수의 종류가 최대 12개라고 가정.
        d = [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]
        
        if v in attr['dic']:
            idx = attr['dic'][v]
        
        else:
            idx = attr['cnt'] 
            attr['dic'][v] = idx
            attr['cnt'] += 1
        
        d[idx] = 1 
        
        ## idx 번째만 1인 리스트가 들어가게 된다.
        exdata.append(d) 
        
    data.append(exdata)
 
data[0]
 
data_train, data_test, label_train, label_test = train_test_split(data, label)