Replacing nan values in one column with mean of same column ( not Null values of column) where other columns have certain values

Question

Replace the missing values in the variable "s_months" and "incidents" by the respective means of the other ships that share the same type AND the same operation period. Here "s_months" and "incidents" are two columns, having nan values, which we want to fill.

I have tried to find means according to asked conditions. But unable to fill nan values in ship data frame. Here are means calculated, and stored as a data frame.

Please show an minimal reproducible example so we can see how far you got. — Joooeey
– Joooeey, Commented Sep 1, 2021 at 7:54
@Joooeey I have calculated means and stored them into another DataFrame. Now, I just have to look at any row or tuple in the first DataFrame where there are nan values. Then check its "type" and "o_period". Take values of means for that "types" and "o_periods" from the second DataFrame and fill in to the first one. — Fateh Muhammad
– Fateh Muhammad, Commented Sep 1, 2021 at 8:06

Corralien · Accepted Answer · 2021-09-01 08:11:31Z

Use groupby and combine_first to fill NaN:

Minimal Reproducible Example:

>>> df types o_periods s_months incidents 0 1 2 63.0 0.0 1 1 2 1095.0 4.0 2 1 2 3353.0 18.0 3 1 2 NaN NaN

keys = ['types', 'o_periods'] vals = ['s_months', 'incidents'] df[vals] = df[vals].combine_first(df.groupby(keys)[vals].transform('mean'))

Output result:

>>> df types o_periods s_months incidents 0 1 2 63.000000 0.000000 1 1 2 1095.000000 4.000000 2 1 2 3353.000000 18.000000 3 1 2 1503.666667 7.333333

Collectives™ on Stack Overflow

Replacing nan values in one column with mean of same column ( not Null values of column) where other columns have certain values

1 Answer 1

Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Related