如何在多个if条件下进行for循环

Question

我有一个df（形状（5928，22）），我正在尝试创建一个新列并根据多个条件添加值。

条件将是：

    if CH == 20 then value = 268,34
    if CH == 24 then value = 322,02
    if CH == 30 then value = 492,65
    if CH == 40 then value = 536,69

    and

    if CH == 20 & ID in (5105561300, 5105561301, 5105561302, 5105561304) then value = 417,43
    if CH == 24 & ID in (5105561300, 5105561301, 5105561302, 5105561304) then value = 500,91
    if CH == 30 & ID in (5105561300, 5105561301, 5105561302, 5105561304) then value = 626,34
    if CH == 40 & ID in (5105561300, 5105561301, 5105561302, 5105561304) then value = 834,85

当我尝试添加新列并根据第一个条件块附加值时，它会很好地工作。

new_value = []

for row in df['CH']:
    if row == 20:
        new_value.append(268.34)
    elif row == 24:
        new_value.append(322.02)
    elif row == 30:
        new_value.append(402.65)
    elif row == 40:
        new_value.append(536.69)
    else:
        new_value.append(0)

df['new_value'] = new_value

当我尝试添加其他条件时，它将无法正常工作。代码类似于：

new_value = []

for row in df['CH']:
    if row == 20 and df['ID'] not in (5105561300, 5105561301, 5105561302, 5105561304):
         new_value.append(268.34)
    elif row == 20 and df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(417.43)
    elif row == 24 and df['ID'] not in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(268.34)
    elif row == 24 and df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(500.91)
    elif row == 30 and df['ID'] not in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(268.34)
    elif row == 30 and df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(626.34)
    elif row == 40 and df['ID'] not in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(268.34)
    elif row == 40 and df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(834.85)
    else:
        new_value.append(0)

    df['new_value'] = new_value

当我尝试上面的代码时，我收到以下错误消息：

  

ValueError：系列的真值不明确。使用a.empty，a.bool（），a.item（），a.any（）或a.all（）。

我不知道该怎么走。在SQL中，我将使用两个简单的WHERE语句，但无法在Python中运行它。

Answer 1

选项1：两个map，一个isin和一个np.where

mtrue  = {20: 268.34, 24: 322.02, 30: 492.65, 40: 536.69}
mfalse = {20: 417.43, 24: 500.91, 30: 626.34, 40: 834.85}
ids = {5105561300, 5105561301, 5105561302, 5105561304}

df['new_value'] = np.where(df['ID'].isin(ids), df['CH'].map(mtrue), df['CH'].map(mfalse))

---

选项2：一个map和一个zip

mtrue  = {20: 268.34, 24: 322.02, 30: 492.65, 40: 536.69}
mfalse = {20: 417.43, 24: 500.91, 30: 626.34, 40: 834.85}
ids = {5105561300, 5105561301, 5105561302, 5105561304}
m = {
    (b, k): v for b, d in zip([True, False], [mtrue, mfalse])
    for k, v in d.items()
}

df['new_value'] = [*map(m.get, zip(df['ID'].isin(ids), df['CH']))]

等效于：

以防万一您可以[*map...]

df['new_value'] = [m[t] for t in zip(df['ID'].isin(ids), df['CH']))]

Answer 2

您的代码问题出在df['ID']中，请更改行循环方式，以解决以下错误消息：

for row, id in zip(df['CH'], df['ID']):
    if row == 20 and id not in (5105561300, 5105561301, 5105561302, 5105561304):
        new_value.append(268.34)
    elif row == 20 and id in (5105561300, 5105561301, 5105561302, 5105561304):
        ...

由于数据集不是很大，因此可以使用列表推导来处理此任务：

# a set of ids to check existence
wlist = { 5105561300, 5105561301, 5105561302, 5105561304 }

# the value of each key is a list with the first element using the value 
# when id not in wlist and the 2nd element the value when id is in wlist
mapping = {
    20: [268.34, 417.43]
  , 24: [322.02, 500.91]
  , 30: [492.65, 626.34]
  , 40: [536.69, 834.85]
}

# new_value will depend on if CH is in mapping and id in wlist
df['new_value'] = [ mapping[ch][int(id in wlist)] if ch in mapping else 0 for ch, id in zip(df.CH, df.ID) ]

Answer 3

您似乎可以将其合并很多，并避免冗余：

default = 268.34

for row in df['CH']:
    id_check = df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304)
    if row == 20:
        new_value = 417.43
    elif row == 24:
        new_value = 500.91
    elif row == 30:
        new_value = 626.34
    elif row == 40
        new_value = 834.85
    else:
        new_value = 0
    df['new_value'] = default if not id_check else value

或者，您可以映射它：

def get_new_value(row):
    d = { 20: 417.43,
             24: 500.91,
             30: 626.34,
             40: 834.85 }
    return d.get(row, 0)

default = 268.34
for row in df['CH']:
    id_check = df['ID'] in (5105561300, 5105561301, 5105561302, 5105561304)
    new_value = default if not id_check else get_new_value(row)

    df['new_value'] = new_value