happy_wealthy的博客

方法一：方法二：

1、pd.concat()df2=pd.DataFrame([[33,44,55]],columns=['a','b','c'])df1 a b c0 1 4 71 2 5 82 3 6 9df2=pd.DataFrame([[33,44,55]],columns=['a','b','c'])df2 a b c0 33 44 55pd.concat([df1,df2],ignore_index=True) a b c0 1 4 71 2 5 8

# 一个空的dfdf = pd.DataFrame(columns=['A'])with open('/Users/zhgi/Desktop/sys_data/android_1k.log') as f: file_lines = f.readlines()for i in file_lines: # 注意这里每次都在复制 df = df.append({'A': i}, ignore_index=True)#方法二：用pandas读取import pandas as p

一、pandas读取数据1、使用pd.read_csv读取数据1.1 读取CSV，使用默认的标题行、逗号分隔符ratings = pd.read_csv(fpath)# 查看前几行数据ratings.head()# 查看数据的形状，返回(行数、列数)ratings.shape# 查看列名列表ratings.columns# 查看索引列ratings.index# 查看每列的数据类型ratings.dtypes1.2 读取txt文件，自己指定分隔符、列名fpath =

pandas缺失值处理data_.loc[:,"Age"] = data_.loc[:,"Age"].fillna(data_.loc[:,"Age"].median())#.fillna 在DataFrame里面直接进行填补

import pymysqlconn = pymysql.connect( host='', user='root', password='y', database='', charset='utf8' )mysql_page = pd.read_sql("select * from ``", con=conn)

df=df[ ~ df['service_name_metric'].str.contains('cpu_0')] #删除某列包含特殊字符的行

pd.Dataframe(rows,columns=[‘hi’])

import pandas as pddf = pd.read_csv('/home/becky/服务器数据处理/meta_info.csv')df.groupby(['company']).count()

time.strptime('1970-01-01 00:25:37.459216','%Y-%m-%d %H:%M:%S.%f')

def hh(q): if q>10: q='大于10' return(q) else: q='小于10' return(q)da["a"]=da["a"].apply(hh)

df_column[column].isnull().any()

df.rename(columns={‘timestamp’:‘time_list’}, inplace = True)

import pandas as pdfrom pandas import Series, DataFrameimport numpy as npdata = DataFrame(np.arange(15).reshape(3,5),index=['one','two','three'],columns=['a','b','c','d','e'])data a b c d e

# 导入实验常用的python包import pandas as pdimport numpy as npimport matplotlib.pyplot as plt# pandas数据写入到csv文件中：names = ['Bob','Jessica','Mary','Mel']births = [968,155,77,973]DataSet = list(zip(names,births))df = pd.DataFrame(data = DataSet ,columns=['Nam