import pandas as pd
train_data = pd.read_csv(filepath_or_buffer='E:\\tmp\\2\\20200818.txt', sep='\t',encoding='utf-8',index_col=0,header=0)
train_data.drop_duplicates(inplace=True)
df = pd.DataFrame(train_data)
# grouped = df.groupby(['通话id'])
df['id']=df.groupby(['通话id']).cumcount()+1
# grouped = grouped.sort_values(['label'], ascending=True)
#
print(df.head(10))
df.to_csv('E:\\tmp\\2\\202008182_2.txt',
sep='\t',
encoding='utf-8',
header=None)