python - Conditional transform on pandas -

i've got simple problem, can't seem right. consider dataframe

df = pd.dataframe({'group' :                 ['a', 'a', 'a', 'b', 'b'], 'time' : [20, 21, 22, 20, 21],                'price' : [3.1, 3.5, 3.0, 2.3, 2.1]})      group price time 0     3.1     20 1     3.5     21 2     3.0     22 3   b   2.3     20 4   b   2.1     21 

now want take standard deviation of price of each group, conditional on being before time 22 (let's call early_std). want create variable information.

the expected result is

group price time    early_std    3.1     20      0.282843   3.5     21      0.282843   3.0     22      0.282843 b   2.3     20      0.141421 b   2.1     21      0.141421 

this tried:

df['early_std'] = df[df.time < 22].groupby('group').\ price.transform(lambda x : x.std()) 

this works gives missing value on time = 22:

 group price    time early_std 0     3.1     20  0.282843 1     3.5     21  0.282843 2     3.0     22  nan 3   b   2.3     20  0.141421 4   b   2.1     21  0.141421 

i tried apply , think works, need reset index, i'd rather avoid (i have large dataset , need repeatedly)

early_std2 = df[df.time < 22].groupby('group').price.std()  df.set_index('group', inplace=true) df['early_std2'] = early_std2      price   time early_std  early_std2 group                  3.1     20  0.282843    0.282843   3.5     21  0.282843    0.282843   3.0     22  nan         0.282843 b   2.3     20  0.141421    0.141421 b   2.1     21  0.141421    0.141421 

thanks!

it looks need add fillna() first code expand std values:

df['early_std'] = df[df.time < 22].groupby('group')['price'].transform(pd.series.std) df['early_std'] = df.groupby('group')['early_std'].apply(lambda x: x.fillna(x.max())) df 

to get:

  group  price  time  early_std 0        3.1    20      0.283 1        3.5    21      0.283 2        3.0    22      0.283 3     b    2.3    20      0.141 4     b    2.1    21      0.141 

edit: have changed ffill more general fillna, use chained .bfill().ffill() achieve same result.