Skip to main content
 首页 » 编程设计

python之如何解析 pandas Dataframe 对象

2024年12月31日9EasonJim

我在 pandas Dataframe 中读取 csv 文件,然后得到它的虚拟对象并连接它们,但是例如我有一个名为“流派”的列,它包含“喜剧、戏剧”和“ Action 、喜剧”,所以当我得到虚拟对象并连接它们时它们为每个句子创建一个对象,但我想解析它们。例如,我想创建对象 'Genre.comedy' , 'Genre.Drama', 'Genre.action' 而不是 'Genre.comedy,drama' 和'Genre.action,comedy' 这是我的代码:

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import csv 
from sklearn import preprocessing 
trainset = pd.read_csv("/Users/yada/Downloads/IMDBMovieData.csv", encoding='latin-1') 
X = trainset.drop(['Description', 'Runtime'], axis=1) 
features = ['Genre','Actors'] 
for f in features: 
    X_dummy = pd.get_dummies(X[f], prefix = f) 
    X = X.drop([f], axis = 1) 
    X = pd.concat((X, X_dummy), axis = 1) 

这是我的 csv 文件的某行: csv

请您参考如下方法:

我认为需要str.get_dummies与 add_prefix :

features = ['Genre','Actors'] 
for f in features: 
    X_dummy = X[f].str.get_dummies(', ').add_prefix(f + '.') 
    X = X.drop([f], axis = 1) 
    X = pd.concat((X, X_dummy), axis = 1) 

或者:

trainset = pd.DataFrame({'Description':list('abc'), 
                   'Genre':['comedy, drama','action, comedy','action'], 
                   'Actors':['a, b','a, c','d, a'], 
                   'Runtime':[1,3,5], 
                   'E':[5,3,6], 
                   'F':list('aaa')}) 
 
print (trainset) 
  Description           Genre Actors  Runtime  E  F 
0           a   comedy, drama   a, b        1  5  a 
1           b  action, comedy   a, c        3  3  a 
2           c          action   d, a        5  6  a 
 
X = trainset.drop(['Description', 'Runtime'], axis=1) 
features = ['Genre','Actors'] 
X_dummy_list = [X.pop(f).str.get_dummies(', ').add_prefix(f + '.') for f in features] 
X = pd.concat([X] + X_dummy_list , axis = 1) 
print (X) 
 
   E  F  Genre.action  Genre.comedy  Genre.drama  Actors.a  Actors.b  \ 
0  5  a             0             1            1         1         1    
1  3  a             1             1            0         1         0    
2  6  a             1             0            0         1         0    
 
   Actors.c  Actors.d   
0         0         0   
1         1         0   
2         0         1