python - Pandas:使用条件和运算符(或和)的 Lambda 函数

Question

我正在理解与 pandas 相关的 lambda 函数。 我正在尝试根据某个单词是否出现在 (A) 列(“上诉”、“审核”、“评估”)中创建一个新列

但是我收到以下错误:

属性错误:“str”对象没有属性“str”

d = {"A":["Assessment Division","Division of Appeals","Calgary Audit","Halifax Market","Georgian Markers"],
"B":[100,300,400,500,400]

sample = pd.DataFrame(d)

sample["results"] = sample["A"].apply(lambda x: "Invoice" if x.str.contains("Audit") or x.str.contains("Appeals") or x.str.contains("Assessment") else "NoAction")

我也尝试过这样做来尝试缩短我的代码，但收到相同的错误:

d = {"A":["Assessment Division","Division of Appeals","Calgary Audit","Halifax Market","Georgian Markers"],
"B":[100,300,400,500,400]

sample = pd.DataFrame(d)

reqs = ["Audit","Appeals","Assessment"]

sample["results"] = sample["A"].apply(lambda x: "Invoice" if x.str.contains().isin(reqs) else "NoAction")

Answer 1

您可以使用 any() 检查字符串是否包含 reqs 中的任何单词:

reqs = ["Audit","Appeals","Assessment"]

mask = sample['A'].apply(lambda x: any(r in x for r in reqs))
sample['results'] = np.where(mask, 'Invoice', 'NoAction')

print(sample)

打印:

                     A    B   results
0  Assessment Division  100   Invoice
1  Division of Appeals  300   Invoice
2        Calgary Audit  400   Invoice
3       Halifax Market  500  NoAction
4     Georgian Markers  400  NoAction

---

编辑:一行:

sample['results'] = sample['A'].apply(lambda x: 'Invoice' if any(r in x for r in reqs) else 'NoAction')