Pandas common methods

DataFrame Get column index name df.columns Get row index name df.index Redefine the column index df.columns=[＂X＂,＂Y＂,＂Z＂] Change the column name of a column again ( If there is a lack of inplace Options , Will not change , Instead, add new columns .)df.rename(columns={'x'：'X'},inplace=True) Print DataFrame Attribute information df.info() see DataFrame Data information of the first five lines df.head() see DataFrame The last five lines of data information df.tail() Create an empty data frame , Easy to add data later

df_empty = pd.DataFrame(columns=['x', 'y', 'z'])

Add data to an empty data box

df_new = pd.DataFrame(np.arange(6).reshape(2, 3), columns=['x', 'y', 'z'])

print(pd.concat([df_empty, df_new], axis=0))

Add a new row to the existing data frame / Column , Or horizontal / Add another table vertically （axis Indicates along the longitudinal axis （axis=0） Or horizontal axis （axis=1） Directional connection ）

dat_cbind = pd.concat([df1, df2], axis=1)

dat_rbind = pd.concat([df1, df3], axis=0)

dat_rbind = df1.append(df3)

Delete rows or columns （ among ,labels Is the row to delete / Column name , Give... With a list ;axis The default is 0, Refers to the deletion of a line , So delete columns When it comes to Appoint axis=1;index Directly specify the row to delete ;columns Directly specify the column to be deleted ;inplace=False, By default, the deletion operation is not Change the original data , Instead, it returns a new... After the deletion operation DataFrame;inplace=True, The original data will be deleted directly operation , And cannot return after deletion ）

df1.drop(labels='y', axis=1, inplace=True)

apply The object of operation is DataFrame A column of （axis=0） Or a certain line （axis=1）applymap The operation object is element level , Act on each DataFrame Every data of map The operation object is also element level , But it is right Series Call the function once for each data in the Sum up by line df_rowsum = df[['2020', '2021']].apply(lambda x: x.sum(), axis=1) Sum by column

df_colsum = df[['2020', '2021']].apply(lambda x: x.sum(), axis=0)