1.選擇DataFrame里面某一列等于某個(gè)值的所有行幽崩，用一條命令即可解決即：
df.loc[df['columnName']=='the value']
2.對某一列的字段值進(jìn)行去重
task_id_sets = df['taskid'].drop_duplicates()
3.Pandas把dataframe轉(zhuǎn)成array
df=df.values
4.對某一列的值出現(xiàn)的次數(shù)進(jìn)行統(tǒng)計(jì)【默認(rèn)情況第一列為索引列】
task_id_all_data['tac_photo'].value_counts()
5..對某一列的值出現(xiàn)的次數(shù)進(jìn)行統(tǒng)計(jì)【對第一列和計(jì)數(shù)列進(jìn)行列名的重命名】
tac_photo_times=task_id_all_data['tac_photo'].value_counts().rename_axis('tac_photo').reset_index(name='counts')
屏幕快照 2020-03-23 下午1.19.58.png
6.將指定列的數(shù)據(jù)信息挑選出來
df_selected = df[['doh_dt','taskcode','tachograph_single_info','taskid','tac_photo']]
7.創(chuàng)建一個(gè)空的dataframe
df = pd.DataFrame(columns = ["ebayno", "p_sku", "sale", "sku"]) #創(chuàng)建一個(gè)空的dataframe
8.指定列名
tac_photo_split.columns=['http','pic','date','ID','num','random_value','jpg']
9.索引——>列
df['index'] = df.index
10.指定行的值
task_id_all_data.loc[[0]]
11.指定行列的值
task_id_all_data.iloc[0,5]
12.排序
task_uuid_all_data.sort_values(by=["tac_time1"],inplace=True,ascending=[True])
13.兩列文本合并成一列
merge_df['type_action'] = merge_df['type'] +"(" + merge_df['action']+")"
14.group-by
a=merge_df.groupby(["index", "type"], as_index=False)['task_uuid'].count()
15.整數(shù)列轉(zhuǎn)換為字符串
merge_df['index'].apply(str)
