pandas 将多个循环优化为一个

Question

我有多个具有相同列的 dfs。这是所有dfs的列表

dfs = [df_14, df_15, df_16, df_17]

例如，每个数据帧看起来像这样，df_14：

id	Days
001	0
004	56
013	95
015	33

接下来，df_15：

Id	Days
001	0
023	18
459	19
811	35

df_16：

Id	Days
111	93
114	56
232	0

df_17：

Id	Days
532	120
113	31
065	58
015	2

我的代码：

rows = [['532', 120],['113', 31], ['065', 58],['025', 2]]
for row in rows:
    df_14.loc[len(df_14)] = row
# and so on

任务是附加到每个月的列表 - 有 30-60 天的列表和另一个单独的列表，其中包含 60-100 天的客户 ID。

#The result should be like this:
14_1:  ['004', '015']
14_2:  ['013']
15_1:  ['811']

我尝试在上面使用 f'strings。比如：

abrreviations = ['14', '15','16', '17']
c = ['_1', '_2']
#Have wrote initializing loops like 
m_list=[]
for a in abrreviations:
    for cp in c:
        m_list.append(a+cp)

这个想法是在带有 f'string 或格式的循环中使用缩写。但是不知道怎么做？或者您能提供其他想法吗？

Answer 1

#let first create a list containing all the dataframe's

all_df=[df_jan, df_feb, df_mar, df_apr, df_may, df_jun, df_jul, df_aug, df_sep, df_oct, df_nov, df_dec]

#create 2 lists for storing the id values of 30-60 range and 90-120 range

list_30,list_90=[],[]

#1 nested for loop for handling all data frames

for cur_df in all_df:
    for id,days in zip(cur_df['Id'],cur_df['Days']):
        if(30<=days<=60):
            list_30.append(id)
        elif(90<=days<=120):
            list_90.append(id)

#Now list_30 and list_90 contains the corresponding id values in that range

希望答案有帮助:)

Answer 2

由于您没有提供数据，所以我做了一个基本示例，它对我有用，所以这里是您描述的单个 for 循环：

import numpy as np
import pandas as pd
dfs = [df_jan, df_feb, df_mar, df_apr, df_may, df_jun, df_jul, df_aug, df_sep, df_oct, df_nov, df_dec]
df30 = []
df90 = []
dfsChained30 = []
dfsChained90 = []
for rowsForMonths, xForMonths in enumerate(dfs):
  # If January [don't consider chain];
  if rowsForMonths == 0:
    for dayN in range(dfs[rowsForMonths]):
      if dfs[rowsForMonths][dayN] in range(30, 61):
        df30.append(dfs[rowsForMonths][dayN])
      elif dfs[rowsForMonths][dayN] in range(90, 121):
        df90.append(dfs[rowsForMonths][dayN])
      else:
        pass
    dfsChained30.append(df30)
    dfsChained90.append(df90)
  # If not January [consider chain];
  else:
    for dayN in range(dfs[rowsForMonths]):
      if dfs[rowsForMonths][dayN] in range(30, 61) and dfs[rowsForMonths][dayN] not in set(dfsChained30):
        df30.append(dfs[rowsForMonths][dayN])
      elif dfs[rowsForMonths][dayN] in range(90, 121) and dfs[rowsForMonths][dayN] not in set(dfsChained90):
        df90.append(dfs[rowsForMonths][dayN])
      else:
        pass
    dfsChained30.append(df30)
    dfsChained90.append(df90)

Answer 3

这可以帮助你

import pandas as pd

data = {'df_jan' : [['001', 0],['004', 56], ['013', 95],['015', 33]],
        'df_feb' : [['001', 0],['023', 18], ['459', 19],['811', 35]],
        'df_mar' : [['111', 93],['114', 56], ['232', 0]],
        'df_apr' : [['532', 120],['113', 31], ['065', 58],['025', 2]]}

dfs = {}
for df in data:
    dfs[df] = pd.DataFrame(data[df], columns=['id', 'days'])

months = {}

for df in dfs:
    months[df.replace('df_', '') + '_30'] = dfs[df][(dfs[df].days >= 30) & (dfs[df].days <= 60)].id.to_list()
    months[df.replace('df_', '') + '_90'] = dfs[df][(dfs[df].days >= 90) & (dfs[df].days <= 120)].id.to_list()
    
months

{'jan_30': ['004', '015'],
 'jan_90': ['013'],
 'feb_30': ['811'],
 'feb_90': [],
 'mar_30': ['114'],
 'mar_90': ['111'],
 'apr_30': ['113', '065'],
 'apr_90': ['532']}

回应您的评论：

我在字典中创建了 df 来简化测试数据的创建。

您的代码可以以自己的方式创建 df ...

df_jan = ...
df_feb = ...
df_mar = ...
df_apr = ...

要处理它们，您需要创建字典 ...

dfs = {
    'df_jan' : df_jan,
    'df_feb' : df_feb,
    'df_mar' : df_mar,
    'df_apr' : df_apr
}

运行循环

您可以将结果分配给您的变量

并删除字典

jan_30 = months['jan_30']
jan_90 = months['jan_90']
feb_30 = months['feb_30']
feb_90 = months['feb_90']
mar_30 = months['mar_30']
mar_90 = months['mar_90']
apr_30 = months['apr_30']
apr_90 = months['apr_90']

del dfs, months