将列与 pandas_schema 进行比较

Question

我正在使用 python 3.8 和 Pandas_schema 对数据运行完整性检查。我要求workflow_next_step 永远不能与workflow_entry_step 相同。我正在尝试生成一个比较两列的 CustomSeriesValidation，因为我没有看到执行此操作的股票函数。

有没有办法使用 Pandas_Schema 比较同一行中的两个单元格值？在此示例中，Pandas_Schema 将为 Mary 返回错误，因为她已从 In Progress 移动到 In Progress。

df = config.pd.DataFrame({
'prospect': ['Bob', 'Jill', 'Steve', 'Mary'], 
'value': [10000, 15000, 500, 50000],
'workflow_entry_step': ['New', 'In Progress', 'Closed', 'In Progress'], 
'workflow_next_step': ['In Progress', 'Closed' ,None, 'In Progress']})

schema = Schema([
Column('prospect', [LeadingWhitespaceValidation(), TrailingWhitespaceValidation()]),
Column('value', [CanConvertValidation(int),'Doesn\'t convert to integer.']),
Column('workflow_entry_step', [InListValidation([None,'New','In Progress','Closed'])]),
Column('workflow_next_step', [CustomSeriesValidation(lambda x: x != Column('workflow_entry_step'), InListValidation([None,'New','In Progress','Closed'])]), 'Steps cannot be the same.')])

Answer 1

import pandas as pd

df = pd.DataFrame({
'prospect': ['Bob', 'Jill', 'Steve', 'Mary'], 
'value': [10000, 15000, 500, 50000],
'workflow_entry_step': ['New', 'In Progress', 'Closed', 'In Progress'], 
'workflow_next_step': ['In Progress', 'Closed' ,None, 'In Progress']})

schema = Schema([
Column('prospect', [LeadingWhitespaceValidation(), TrailingWhitespaceValidation()]),
Column('value', [CanConvertValidation(float)]),
Column('workflow_entry_step', [InListValidation([None,'New','In 
Progress','Closed'])]),
Column('workflow_next_step', [CustomSeriesValidation(lambda x: x != 
df['workflow_entry_step'], 'Steps cannot be the same.'), 
InListValidation([None,'New','In Progress','Closed'])])
])
errors = schema.validate(df)
for error in errors:
    print(error)

输出：

{row: 3, column: "workflow_next_step"}: "In Progress" Steps cannot be the same.