In [1]:

    

import pandas as pd


    

In [2]:

    

df = pd.read_csv('data/src/sample_pandas_normal.csv')
s = df['state']


    

In [3]:

    

%timeit s.map({'NY': 'NewYork', 'CA': 'California', 'TX': 'Texas'})


    

    




345 µs ± 14 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [4]:

    

%timeit s.replace({'NY': 'NewYork', 'CA': 'California', 'TX': 'Texas'})


    

    




519 µs ± 15.9 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [5]:

    

s_copy = s.copy()
%timeit s_copy.update(s_copy.map({'NY': 'NewYork'}))


    

    




643 µs ± 21.4 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [6]:

    

s_copy = s.copy()
%timeit s_copy.replace({'NY': 'NewYork'}, inplace=True)


    

    




230 µs ± 10 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [7]:

    

s_copy = s.copy()
%timeit s_copy.update(s_copy.map({'NY': 'NewYork', 'CA': 'California', 'TX': 'Texas'}))


    

    




627 µs ± 10.9 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [8]:

    

s_copy = s.copy()
%timeit s_copy.replace({'NY': 'NewYork', 'CA': 'California', 'TX': 'Texas'}, inplace=True)


    

    




441 µs ± 22.1 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [9]:

    

%timeit s.map({'NY': 'NewYork'})


    

    




335 µs ± 7.55 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [10]:

    

%timeit s.replace({'NY': 'NewYork'})


    

    




261 µs ± 6.57 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [11]:

    

df = pd.read_csv('data/src/titanic_train.csv')
s = df['Sex']


    

In [12]:

    

%timeit s.map({'male': 0, 'female': 1})


    

    




422 µs ± 30.6 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)

In [13]:

    

%timeit s.replace({'male': 0, 'female': 1})


    

    




634 µs ± 13.4 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)