迭代Python中的行

Question

我想首先遍历项目编号，然后对于特定项目，我需要遍历周数。主要目的是标记每个项目的周数，其中有库存添加。我是新手，所以我不知道如何进行双重迭代。

输出将是这样的。当前周的库存大于前一个库存时，该标志变为1。

Answer 1

IIUC：您可以尝试使用.groupby项目编号，然后使用.shift中的dataframe，如下所示。

假设原始数据帧如下：

   DIV_NBR  itm_nbr  WEEK_NO  DISTINCT_ITM_CNT  INVENTORY_IN_HAND
0       18    65874   201511               5.0             2925.0
1       18    65874   201512               5.0             2910.0
2       18    65874   201513               5.0             2961.0
3       19    65875   201511               5.0             2965.0
4       19    65875   201512               5.0             2971.0

然后：

# keep record of last week by grouping by item number and then using shift
df['LAST_WEEK'] = df.groupby('itm_nbr')['INVENTORY_IN_HAND'].shift()

# check if current inventory is greater than last week
df['Flag'] = (df['INVENTORY_IN_HAND'] - df['LAST_WEEK'])>0

# delete additional column
del df['LAST_WEEK']

# change flag int
df['Flag'] = df['Flag'].astype(int)

print(df)

结果：

   DIV_NBR  itm_nbr  WEEK_NO  DISTINCT_ITM_CNT  INVENTORY_IN_HAND  Flag
0       18    65874   201511               5.0             2925.0     0
1       18    65874   201512               5.0             2910.0     0
2       18    65874   201513               5.0             2961.0     1
3       19    65875   201511               5.0             2965.0     0
4       19    65875   201512               5.0             2971.0     1