I have a pandas dataframe which is indexed by strings. Let's say my index looks like df.index = ['AA','AB','AC',...] and I want to access df.loc['AC':'AE'], which works well.
Is there any way to get the position of these indices, giving me ['AC':'AE'] => [2,3,4]? I know there is df.index.get_loc('AC') => 2 but this works only for single values and not for lists.
Use:
df = pd.DataFrame({'a': [5,6,7,8, 10]}, index=['AA','AB','AC','AD','AE'])
pos = list(range(df.index.get_loc('AC'), df.index.get_loc('AE') + 1))
print (pos)
[2, 3, 4]
Another solutions with Index.searchsorted:
pos = list(range(df.index.searchsorted('AC'), df.index.searchsorted('AE') + 1))
print (pos)
[2, 3, 4]
a = df.index.searchsorted(['AC', 'AE'])
pos = list(range(a[0], a[1] + 1))
print (pos)
[2, 3, 4]
You can define a function to extract the integer range:
df = pd.DataFrame(np.arange(7), index=['AA','AB','AC','AD','AE','AF','AG'])
def return_index(df, a, b):
col_map = df.index.get_loc
return np.arange(col_map(a), col_map(b)+1)
res = return_index(df, 'AC', 'AE')
print(res)
array([2, 3, 4])
