python - 从数据框中获取作为日期时间对象的索引(Get index as datetime object from dataframe)

Question

I got a csv file that looks like this

(我有一个看起来像这样的csv文件)

valid,temp,pressure
2019-11-25 12:00,22,4
2019-11-22 12:00,24,4
2019-11-20 12:00,24,5
2019-11-17 12:00,26,5

I read the csv as a pandas dataframe and set the valid column as the index.

(我将csv读取为pandas数据框，并将有效列设置为索引。)

I also convert it to a pd.datetime

(我也将其转换为pd.datetime)

df = pd.read_csv(file)
pd.to_datetime(df['valid'])  # convert 'valid' column to pd.datetime objects
df = df.set_index('valid')  # set the 'valid' column as index

Now I want to get the index as a datetime object of a certain row.

(现在，我想将索引作为某一行的日期时间对象。)

How do I do that?

(我怎么做？)

I tried this but it doesn't work

(我试过了但是没用)

row = df.iloc[2]
print(row.index.month)  # using .month just see if the returned object is a pd.datetime

AttributeError: 'Index' object has no attribute 'month'

  ask by Vader translate from so

Answer 1

You forgot to assign back, because to_datetime is not inplace function and then test Series.name , because if select one row it return Series with name by index of row in DataFrame :

(您忘记分配回去，因为to_datetime不是inplace函数，然后测试Series.name ，因为如果选择一行，则返回带有名称的Series （按DataFrame的行索引）：)

df = pd.read_csv(file)
df['valid'] = pd.to_datetime(df['valid'])
df = df.set_index('valid')
row = df.iloc[2]

print (row)
temp        24
pressure     5
Name: 2019-11-20 12:00:00, dtype: int64

print (row.name)
2019-11-20 12:00:00

print (row.name.month)
11

Also for convert column to datetimes is possible use parse_dates and index_col parameter in read_csv for DatetimeIndex :

(对于将列转换为日期时间，也可以在read_csv为DatetimeIndex使用parse_dates和index_col参数：)

df = pd.read_csv(file, parse_dates=['valid'], index_col=['valid'])
row = df.iloc[2]
print (row.name.month)
11