pandas操作

• pandas知识回顾
• iloc/ix切片
• 列条件与与列的筛选
• 读取csv、xlsx文件
• 行的增加与删除
• 列的增加与删除
• 排序
• 数据分组描述
• 统计描述
• 作图
• 数据框合并

#-*-encoding:utf-8-*-
'''
created by zwg in 2016-12-03
'''
import pandas
import numpy
from pylab import mpl
from matplotlib import pyplot
mpl.rcParams['font.sans-serif']=['SimHei']
mpl.rcParams['axes.unicode_minus']=False

def practise_one():
    #pandas知识回顾
    numpy.random.seed()
    Data = pandas.DataFrame(data=numpy.random.randn(5, 3), columns=list('ABC'))
    # print Data
    # print Data.values
    # print Data.columns
    # print Data.index
    # print Data.tail(5)
    # print Data.head(5)

    #iloc/ix切片
    print Data.iloc[0:2, [0, 2]]
    print Data.ix[0:2, [0, 2]]
    print Data.ix[0:2, ['A', 'C']]
    print Data.iat[0, 0]

    # 列条件与与列的筛选
    print Data[Data.A > 0.5]
    print Data[(Data.A > 0.5) & (Data.C < 0.7)]
    print Data[['B']]

    print Data[Data.A > 0.5].ix[:, 'B']
    print Data[Data.A > 0.5][['B']]
    print Data[['B']][Data.A > 0.5]

    #读取csv、xlsx文件
    # pandas.read_csv(filepath_or_buffer=file_name,header=0/None,index_col=Fasle/row_name,encoding='utf-8')
    # pandas.read_table(filepath_or_buffer=file_name,header=0/None,index_col=Fasle/row_name,encoding='utf-8')
    # pandas.read_excel(io=file_name,sheetname=0/sheetname,header=0/None,index_col=Fasle/row_name,encoding='utf-8')


def practise_two():
    Data=pandas.read_excel('test.xlsx',encoding='utf-8',header=0)

    # 行的增加与删除
    # Data1 = Data.drop([1,2,3], axis=0)
    # Data2=Data1.copy()
    # Data1=Data1.append(Data2,ignore_index=True)
    # print Data1

    # 列的增加与删除
    # Data1=Data.drop(['name', 'class'],axis=1)
    # print Data1.columns
    # Data1=Data.reindex(columns=['class','name','grade','add'])
    # print Data1.columns

    #排序
    # print Data.sort_values(by=['grade'],ascending=False)
    # print Data.sort_index(axis=0,ascending=False)

    # 数据分组描述
    # Data1=Data.groupby('sex')
    # print Data1['sex'].count()
    # print Data1['grade'].mean()
    # print Data1['sex'].unique()

    # 数据分组
    # Data2=Data.groupby(['class','sex'])
    # print Data2['grade'].describe()

    # 统计描述
    # print Data.describe(include='all')
    # Data3=Data[['grade','sex']]
    # figure=pyplot.figure()
    # Data3.plot(kind='box',by='sex')
    # pyplot.show()

    #分组计数
    # print Data['sex'].value_counts()

    # 作图
    # Data.boxplot(column='grade',by=['class','sex'])
    # Data.hist(column='grade')
    # Data['sex'].value_counts().plot(kind='bar')
    Data.groupby(['class'])['grade'].mean().plot(kind='barh',colormap='cool')
    Data.plot(x='grade',y='age',kind='scatter',title='grade-age change',logx=True,logy=True)
    Data.plot(kind='kde')
    pyplot.show()

def practise_three():
    #数据框合并
    Data1=pandas.DataFrame(data=numpy.random.rand(5,3),columns=list('ABC'))
    Data1['D'] = [1, 2, 3, 4, 5]
    print Data1
    Data2 = pandas.DataFrame(data=numpy.random.rand(5, 3), columns=list('ABC'))
    Data2['D'] = [1, 2, 2, 4, 5]
    print Data2
    # join数据框合并和merge数据框合并
    # Data1=Data1.set_index('D')
    # Data2=Data2.set_index('D')
    # Data3=Data1.join(Data2,lsuffix='_left',rsuffix='_right',how='left')
    # print Data3

    Data4=Data1.merge(Data2,on='D',how='inner',suffixes=('_1','_2'))
    print Data4




if __name__=='__main__':
    # practise_one()
    practise_two()
    # practise_three()