Day 08. MapFunc_OneHotEncode_Join

# -*- coding: utf-8 -*-
"""Day 08_Mission_baby_Movie.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1XwL5qAxvCR403REB8dT2o2W3PCjrYL5u
"""
 
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
 
@interact_manual( x =IntSlider(0,0,12) )
def test_model( x ):
    print(x)
 
years = range( 1880, 2011 )
 
pieces = []
for year in years:
    
    path  = "/gdrive/My Drive/Colab Notebooks/resources/babynames/yob%d.txt" % year
    frame = pd.read_csv( path, names = ['name','gender','birth'] ) 
    frame['year'] = year
    
    pieces.append( frame )
    
len(pieces)
 
df = pd.concat(pieces, ignore_index = True)
 
df[:2]
 
## 1880년 부터 2010년 사이에 이름에 들어간 알파벳의 빈도수
alphabet_dic = {}
 
if 'a' not in alphabet_dic:
    alphabet_dic['a'] = 1
else :
    alphabet_dic['a'] += 1
    
alphabet_dic
 
df['name'].size  # 1690784
 
def acount_alpha( name ):
    for c in name:
 
        if c not in alphabet_dic:
            alphabet_dic[c]  = 1
        else :
            alphabet_dic[c] += 1
 
 
 
df['name'].map(acount_alpha)
 
import operator
sorted_x = sorted(alphabet_dic.items(), key=operator.itemgetter(1) , reverse = True)
 
## 상위 10 개
sorted_x[:10]
 
# [('a', 1425321),
#  ('e', 1193584),
#  ('n', 905036),
#  ('i', 828887),
#  ('r', 688878),
#  ('l', 669406),
#  ('o', 469078),
#  ('t', 337296),
#  ('s', 333697),
#  ('y', 321302)]
 
"""#Movielens"""
 
movies   = pd.read_csv("/gdrive/My Drive/Colab Notebooks/resources/movielens/movies.dat", header = None , sep = "::" , names = ['movie_id', 'title','genres'  ])
 
movies['year'] = movies['title'].str.extract(" \((\d{4})\)")
 
movies['title'] = movies['title'].str.replace(" \((\d{4})\)", "")
 
## 장르 하나씩만 추출해서 원핫 인코딩
movies['genres'] = movies['genres'].str.replace("\|.*", "")
 
## 장르 따로 저장
genres_names = movies['genres'].value_counts().index
 
genres_names = np.array(genres_names).reshape(-1,1)
 
from sklearn.preprocessing import OneHotEncoder
ohe = OneHotEncoder(n_values = "auto", handle_unknown = "ignore")
ohe.fit( genres_names )
 
movies['genres'] = pd.Series(list(ohe.transform(movies[['genres']]).toarray()))
movies[:10]
 
#################################################################
####### 피처 이름 변경 및 인덱스 타입 변경 조인 코드 저장용######
#################################################################
 
df = tempDf.rename(columns = { 0: 'genres'})
tempDf = pd.concat([pd.Series(genres_names.flatten()), pd.DataFrame(tempOh)], axis = 1 ,ignore_index = True)
movies.columns
movies.join(df.set_index('genres'), on='genres')[:2]