Python Pandas to_datetime ValueError：未知的字符串格式
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/34505579/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Pandas to_datetime ValueError: Unknown string format
提问by pheno
I have a column in my (pandas) dataframe:
我的（熊猫）数据框中有一列：
data['Start Date'].head()
type(data['Start Date'])
Output:
1/7/13
1/7/13
1/7/13
16/7/13
16/7/13
<class 'pandas.core.series.Series'>
When I convert it into a date format (as follows) I am getting the error ValueError: Unknown string format
当我将其转换为日期格式（如下）时，出现错误ValueError: Unknown string format
data['Start Date']= pd.to_datetime(data['Start Date'],dayfirst=True)
...
...
/Library/Python/2.7/site-packages/pandas/tseries/tools.pyc in _convert_listlike(arg, box, format, name)
381 return DatetimeIndex._simple_new(values, name=name, tz=tz)
382 except (ValueError, TypeError):
--> 383 raise e
384
385 if arg is None:
ValueError: Unknown string format
What am I missing here?
我在这里缺少什么？
采纳答案by jezrael
I think the problem is in data - a problematic string exists. So you can try check length of the string in column Start Date:
我认为问题出在数据中 - 存在有问题的字符串。所以你可以尝试检查列中字符串的长度Start Date：
import pandas as pd
import io
temp=u"""Start Date
1/7/13
1/7/1
1/7/13 12 17
16/7/13
16/7/13"""
data = pd.read_csv(io.StringIO(temp), sep=";", parse_dates=False)
#data['Start Date']= pd.to_datetime(data['Start Date'],dayfirst=True)
print data
Start Date
0 1/7/13
1 1/7/1
2 1/7/13 12 17
3 16/7/13
4 16/7/13
#check, if length is more as 7
print data[data['Start Date'].str.len() > 7]
Start Date
2 1/7/13 12 17
Or you can try to find these problematic row different way e.g. read only part of the datetime and check parsing datetime:
或者您可以尝试以不同的方式找到这些有问题的行，例如仅读取日期时间的一部分并检查解析日期时间：
#read first 3 rows
data= data.iloc[:3]
data['Start Date']= pd.to_datetime(data['Start Date'],dayfirst=True)
But this is only tips.
但这只是提示。
EDIT:
编辑：
Thanks joris for suggestion add parameter errors ='coerce'to to_datetime:
感谢 joris 的建议将参数添加errors ='coerce'到to_datetime：
temp=u"""Start Date
1/7/13
1/7/1
1/7/13 12 17
16/7/13
16/7/13 12 04"""
data = pd.read_csv(io.StringIO(temp), sep=";")
#add parameter errors coerce
data['Start Date']= pd.to_datetime(data['Start Date'], dayfirst=True, errors='coerce')
print data
Start Date
0 2013-07-01
1 2001-07-01
2 NaT
3 2013-07-16
4 NaT
#index of data with null - NaT to variable idx
idx = data[data['Start Date'].isnull()].index
print idx
Int64Index([2, 4], dtype='int64')
#read csv again
data = pd.read_csv(io.StringIO(temp), sep=";")
#find problematic rows, where datetime is not parsed
print data.iloc[idx]
Start Date
2 1/7/13 12 17
4 16/7/13 12 04
