根据熊猫数据框中的多行添加列

Question

如何根据对另一个数据框多行中的值的操作在数据框中添加一列？

所以这是我最初的数据框示例。

我想要输出如下

在哪里

例子

到目前为止，我尝试使用 unique(ord_date、crt_code 和 del_date 组合) 添加一个新数据框，然后尝试计算每一行的分数，但我不知道如何设置 if 条件。

df2['score'][(df2['ord_date']==xxxx)&(df2['crt_code']==xxxx)&(df2['del_date']==xxxx)] 

= if(df['val1'][(df['slb_qty']==2)&(df['ord_date']==xxxx)&(df['crt_code']==xxxx)&(df['del_date']==xxxx)] + df['val1'][(df['slb_qty']==12)&(df['ord_date']==xxxx)&(df['crt_code']==xxxx)&(df['del_date']==xxxx)] >=80 ) then 200

加上这将成为一个非常大的语句来检查所有 4 个难以阅读的条件。

谁能建议如何以更清洁/简单的方式实现我想要的输出？

Answer 1

你需要收集独特的价值每个唯一值的总和数量为他们计算分数下次将数据作为文本而不是图像发布。我的代码与描述：=^..^=import pandas as pdfrom io import StringIOdata = StringIO("""ord_date crt_code del_date slb_qty val101/01/2019 125 10/01/2019 2 3801/01/2019 125 10/01/2019 4 2701/01/2019 125 10/01/2019 12 3501/01/2019 128 10/01/2019 2 4501/01/2019 128 10/01/2019 4 2101/01/2019 128 10/01/2019 12 2301/01/2019 128 10/01/2019 14 2402/01/2019 125 10/01/2019 2 3702/01/2019 125 10/01/2019 12 3002/01/2019 125 10/01/2019 4 2902/01/2019 128 10/01/2019 14 2202/01/2019 128 10/01/2019 4 2602/01/2019 128 10/01/2019 12 2102/01/2019 128 10/01/2019 2 29""")# load datadf = pd.read_csv(data, sep=" ")# get unique valuesdf_unique = df.groupby(['ord_date', 'crt_code', 'del_date']).size().reset_index()# drop last columndf_unique = df_unique.drop([0], axis=1)# sum quantity valuesslb_qty_2_12 = []slb_qty_4_14 = []for index, row in df_unique.iterrows():    # select row range from raw data    selected_rows = df[(df['ord_date'] == row['ord_date']) & (df['crt_code'] == row['crt_code']) & (df['del_date'] == row['del_date'])]    # find 2 and 12 qty    rows_2_12 = selected_rows[(selected_rows['slb_qty'] == 2) | (selected_rows['slb_qty'] == 12)]    # sum values    values_sum = rows_2_12['val1'].sum()    # collect data    slb_qty_2_12.append(values_sum)    # find 4 and 14 qty    rows_4_14 = selected_rows[(selected_rows['slb_qty'] == 4) | (selected_rows['slb_qty'] == 14)]    # sum values    values_sum = rows_4_14['val1'].sum()    # collect data    slb_qty_4_14.append(values_sum)# add calculated values to data framedf_unique['slb_qty_2_12'] = slb_qty_2_12df_unique['slb_qty_4_14'] = slb_qty_4_14# calculate scorescore = []for index, row in df_unique.iterrows():    if row['slb_qty_4_14'] >= 80:        score.append(300)    elif 80 > row['slb_qty_4_14'] >= 60:        score.append(150)    elif row['slb_qty_2_12'] >= 80:        score.append(200)    elif 80 > row['slb_qty_2_12'] >= 60:        score.append(100)    else:        score.append(0)# drop used columnsdf_unique = df_unique.drop(['slb_qty_2_12', 'slb_qty_4_14'], axis=1)# add scoredf_unique['Score'] = score输出：     ord_date  crt_code    del_date  Score0  01/01/2019       125  10/01/2019    1001  01/01/2019       128  10/01/2019    1002  02/01/2019       125  10/01/2019    1003  02/01/2019       128  10/01/2019      0