Pandas / Python - 数据框和字典之间的多个条件匹配

Question

我有 Pandas 数据框和一个字典，每个键有两个值。如何检查字典中的两个值与数据框中的两个列值之间的匹配？如果匹配，我需要返回匹配的字典元素的键和第三个数据框列值。

示例数据框：

MACHINE     MONTH       possible hours

M301        December    0.051136
M304        December    0.215909
M305        December    0.230114
M306        December    0.198864
M307        December    0.130682
M400        November    0.122024
M400        December    0.252841
M714        November    0.005952

字典示例：

mapping_dict = {"O36": ("November", "M101"), "O37": ("November", "M102"), "O38": ("November", "M103"), "O39": ("November", "M104"),
             "O40": ("November", "M105"), "O41": ("November", "M106"), "O42": ("November", "M107"), "O43": ("November", "M201"),
             "O44": ("November", "M202"), "O45": ("November", "M203"), "O46": ("November", "M204"), "O47": ("November", "M205"),
             "O48": ("November", "M206"), "O49": ("November", "M207"), "O50": ("November", "M301"), "O51": ("November", "M302"),
             "P36": ("December", "M101"), "P37": ("December", "M102"), "P38": ("December", "M103"), "P39": ("December", "M104"),
             "P40": ("December", "M105"), "P41": ("December", "M106"), "P42": ("December", "M107"), "P43": ("December", "M201"),
             "P44": ("December", "M202"), "P45": ("December", "M203"), "P46": ("December", "M204"), "P47": ("December", "M205"),
             "P48": ("December", "M206"), "P49": ("December", "M207"), "P50": ("December", "M301"), "P51": ("December", "M302"),
             "P52": ("December", "M303"), "P53": ("December", "M304"), "P54": ("December", "M305"), "P55": ("December", "M306"),
             "P56": ("December", "M307"), "P57": ("December", "M400"), "P58": ("December", "M401"), "P59": ("December", "M402"),
             "P60": ("December", "M403"), "P61": ("December", "M404"), "P62": ("December", "M405"), "P63": ("December", "M406"),
             "P64": ("December", "M407"), "P65": ("December", "M712"), "P66": ("December", "M713"), "P67": ("December", "M714")}

我需要将possible hours 的值写入 Excel 电子表格中的特定单元格。该单元格是字典中的key。必须写入的数据是possible hours 列。

例如，在检查M400 和December 的df 和dict 时，我需要返回0.252841 和P57。如果它有帮助 df 和 dict 中的所有值都是字符串（或对象）。

Answer 1

你可以试试：

df = pd.DataFrame([['M301', 'December', 0.051136], ['M304', 'December', 0.215909],
                   ['M305', 'December', 0.230114], ['M306', 'December', 0.198864],
                   ['M307', 'December', 0.130682], ['M400', 'November', 0.122024],
                   ['M400', 'December', 0.252841], ['M714', 'November', 0.005952]],
                  columns=['MACHINE', 'MONTH', 'possible hours'])

mapping_dict = {"O36": ("November", "M101"), "O37": ("November", "M102"), "O38": ("November", "M103"), "O39": ("November", "M104"),
             "O40": ("November", "M105"), "O41": ("November", "M106"), "O42": ("November", "M107"), "O43": ("November", "M201"),
             "O44": ("November", "M202"), "O45": ("November", "M203"), "O46": ("November", "M204"), "O47": ("November", "M205"),
             "O48": ("November", "M206"), "O49": ("November", "M207"), "O50": ("November", "M301"), "O51": ("November", "M302"),
             "P36": ("December", "M101"), "P37": ("December", "M102"), "P38": ("December", "M103"), "P39": ("December", "M104"),
             "P40": ("December", "M105"), "P41": ("December", "M106"), "P42": ("December", "M107"), "P43": ("December", "M201"),
             "P44": ("December", "M202"), "P45": ("December", "M203"), "P46": ("December", "M204"), "P47": ("December", "M205"),
             "P48": ("December", "M206"), "P49": ("December", "M207"), "P50": ("December", "M301"), "P51": ("December", "M302"),
             "P52": ("December", "M303"), "P53": ("December", "M304"), "P54": ("December", "M305"), "P55": ("December", "M306"),
             "P56": ("December", "M307"), "P57": ("December", "M400"), "P58": ("December", "M401"), "P59": ("December", "M402"),
             "P60": ("December", "M403"), "P61": ("December", "M404"), "P62": ("December", "M405"), "P63": ("December", "M406"),
             "P64": ("December", "M407"), "P65": ("December", "M712"), "P66": ("December", "M713"), "P67": ("December", "M714")}

def find(x, y):
    for k, v in mapping_dict.items():
        if v == (x, y):
            return k

df['New'] = [find(tuple(r[1].values)[1], tuple(r[1].values)[0]) for r in df.iloc[:,:2].iterrows()]

print(df)

结果：

  MACHINE     MONTH  possible hours   New
0    M301  December        0.051136   P50
1    M304  December        0.215909   P53
2    M305  December        0.230114   P54
3    M306  December        0.198864   P55
4    M307  December        0.130682   P56
5    M400  November        0.122024  None
6    M400  December        0.252841   P57
7    M714  November        0.005952  None

Answer 2

与merge 一起使用左连接并将嵌套字典转换为帮助器DataFrame，列名被省略，因为df1 和df 之间的列名交叉合并：

df1 = pd.DataFrame([(k, v1, v2) for k, (v1, v2) in mapping_dict.items()], 
                    columns = ['val','MONTH','MACHINE'])
df = df.merge(df1, how='left')
print (df)
  MACHINE     MONTH  possible hours  val
0    M301  December        0.051136  P50
1    M304  December        0.215909  P53
2    M305  December        0.230114  P54
3    M306  December        0.198864  P55
4    M307  December        0.130682  P56
5    M400  November        0.122024  NaN
6    M400  December        0.252841  P57
7    M714  November        0.005952  NaN

---

df1 = df.merge(df1, how='left').dropna(subset=['val'])[['possible hours','val']]
print (df1)
   possible hours  val
0        0.051136  P50
1        0.215909  P53
2        0.230114  P54
3        0.198864  P55
4        0.130682  P56
6        0.252841  P57

如果需要功能：

def lookup(machine, month):
    c = [k for k,(v1, v2) in mapping_dict.items() if (v1 == month) and (v2 == machine)]
    h = df.loc[df['MACHINE'].eq(machine) & df['MONTH'].eq(month), 'possible hours']
    try:
        return(c[0], h.iat[0])
    except:
        return ('no match')

print (lookup('M400','December'))
('P57', 0.252841)

print (lookup('M4','December'))
no match