Pandas can't select index range as string date-CodePudding

csv.index.values contains the below data

[263 rows x 13 columns]>
['31/01/1997' '28/02/1997' '31/03/1997' '30/04/1997' '31/05/1997'...

csv.loc['01/01/2009':'31/12/2015']

    raise KeyError(key) from err
KeyError: '01/01/2009'

Why can't I select a range?

CodePudding user response：

.loc works with dataframes, but you need to use it in an array.

In other words, csv.loc.index.values must return an array, so that you can use it with .loc : from pandas.DataFrame.loc documentation : Access a group of rows and columns by label(s) or a boolean array.

I would recomend you to first load your csv file, transfer it to dataframe, set the desired column date as index (watch out for inplace = true) and finally apply .loc

Here an example:

# creating some data
new_list= np.arange(20) # creating some data
BlackHoles = np.random.randint(0,2,20)

#converting data to dataframe
df=pd.DataFrame({'date' : new_list, 'BlackHoles' : BlackHoles})
#converting new_list to Timestamp
df['date']=pd.to_numeric(df['date'])
df['date']=pd.to_datetime(df['date'],dayfirst=True,unit='D',origin='15.04.2022')

#setting the desired column as index ( this point is missing in your code)
df_test=df.set_index('date')

df_test.index.values # will return as: array(['.....],],dtype='datetime64[ns]')
df_test.index # will give you the possibilites of "indexes" that can be called
df_test.loc['2022-04-16': '2022-04-25']

Be advised, since you are working with Timestamps, be sure to eliminate duplicates, or you will get the unique errors. In this case I suggest you to use the groupby in your date column.

I hope I could assist on this one.

CodePudding user response：

The index dates were strings instead of datetime objects -- needed to convert.

csv = csv.to_datetime(csv.index)