Request: show datetime percentiles in 'describe' method · Issue #30164 · pandas-dev/pandas (original) (raw)

Would be nice if pandas.Series.describe would show percentiles of datetime dtypes in the same way it does for other numeric types:

import pandas as pd dt_list = [ "2019-01-01 00:01:00", "2019-01-01 00:02:00", "2019-01-01 00:02:10", "2019-01-01 00:10:01", "2019-01-01 00:11:00" ] s = pd.to_datetime(pd.Series(dt_list)) s.describe()

count                       5
unique                      5
top       2019-01-01 00:02:00
freq                        1
first     2019-01-01 00:01:00
last      2019-01-01 00:11:00
dtype: object

s.describe(percentiles = [0.25, 0.5, 0.75])

doesn't show percentiles

Compare to R:

dt <- c( "2019-01-01 00:01:00", "2019-01-01 00:02:00", "2019-01-01 00:02:10", "2019-01-01 00:10:01", "2019-01-01 00:11:00" ) summary(as.POSIXct(dt))

                 Min.               1st Qu.                Median                  Mean               3rd Qu.                  Max. 
"2019-01-01 00:01:00" "2019-01-01 00:02:00" "2019-01-01 00:02:10" "2019-01-01 00:05:14" "2019-01-01 00:10:01" "2019-01-01 00:11:00"