R, Python 분석과 프로그래밍의 친구 (by R Friend)

  (1) DataFrame으로 부터 특정 개수의 표본을 무작위(확률)로 추출하기 (number)

import pandas as pd

df = pd.DataFrame({'num_legs': [2, 4, 8, 0],
                   'num_wings': [2, 0, 0, 0],
                   'num_specimen_seen': [10, 2, 1, 8]},
                  index=['falcon', 'dog', 'spider', 'fish'])

df

df.sample(n=2, # number of items from axis to return.
          random_state=1004) # seed for random number generator for reproducibility

  (2) DataFrame으로 부터 특정 비율의 표본을 무작위로 추출하기 (fraction)

df.sample(frac=0.5, # fraction of axis items to return.
          random_state=1004)

## ValueError: Replace has to be set to `True` when upsampling the population `frac` > 1.
df.sample(frac=1.5,
          random_state=1004)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-45-2fcc4494d7ae> in <module>
----> 1 df.sample(frac=1.5, # fraction of axis items to return. 
      2           random_state=1004)

~/opt/anaconda3/lib/python3.8/site-packages/pandas/core/generic.py in sample(self, n, frac, replace, weights, random_state, axis)
   5326             n = 1
   5327         elif frac is not None and frac > 1 and not replace:
-> 5328             raise ValueError(
   5329                 "Replace has to be set to `True` when "
   5330                 "upsampling the population `frac` > 1."

ValueError: Replace has to be set to `True` when upsampling the population `frac` > 1.

## parameter 'n' and 'frac' cannot be used at the same time.
## ValueError: Please enter a value for `frac` OR `n`, not both
df.sample(n=2, frac=0.5)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-6-b31ebc150882> in <module>
      1 ## parameter 'n' and 'frac' cannot be used at the same time.
      2 ## ValueError: Please enter a value for `frac` OR `n`, not both
----> 3 df.sample(n=2, frac=0.5)

~/opt/anaconda3/lib/python3.8/site-packages/pandas/core/generic.py in sample(self, n, frac, replace, weights, random_state, axis)
   5335             n = int(round(frac * axis_length))
   5336         elif n is not None and frac is not None:
-> 5337             raise ValueError("Please enter a value for `frac` OR `n`, not both")
   5338 
   5339         # Check for negative sizes

ValueError: Please enter a value for `frac` OR `n`, not both

  (3) DataFrame으로 부터 복원 무작위 표본 추출하기

      (random sampling with replacement)

## replace=True: random sampling with replacement
df.sample(n=8, # or equivalently: frac=2
          replace=True, # random sampling with replacement
          random_state=1004)

## ValueError: Cannot take a larger sample than population when 'replace=False'
df.sample(n=8,
          replace=False # random sampling without replacement
)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-42-40c76bd4c271> in <module>
      1 ## replace=True: random sampling with replacement
----> 2 df.sample(n=8, # or equivalently: frac=2
      3           replace=False # random sampling without replacement
      4 )

~/opt/anaconda3/lib/python3.8/site-packages/pandas/core/generic.py in sample(self, n, frac, replace, weights, random_state, axis)
   5343             )
   5344 
-> 5345         locs = rs.choice(axis_length, size=n, replace=replace, p=weights)
   5346         return self.take(locs, axis=axis)
   5347 

mtrand.pyx in numpy.random.mtrand.RandomState.choice()

ValueError: Cannot take a larger sample than population when 'replace=False'

  (4) DataFrame으로 부터 가중치를 부여하여 표본 추출하기 (weights)

## Using a DataFrame column as weights.
## Rows with larger value in the num_specimen_seen column are more likely to be sampled.
df.sample(n=2,
          weights='num_specimen_seen'

)

  (5) DataFrame으로 부터 칼럼에 대해 무작위 표본 추출하기 (axis=1, axis='column)

## Axis to sample: by column
df.sample(n=2,
          random_state=1004,
          axis=1) # or equivalently, axis='column'

## Axis to sample: by index
df.sample(n=2,
          random_state=1004,
          axis=0) # or equivalently, axis='index', default

  (6) DataFrame으로 부터 특정 칼럼에 대해 무작위 표본 추출한 결과를

       numpy array로 할당하기

## Sampling only for a column
df['num_legs'].sample(n=2, random_state=1004)

[Out] 
falcon    2
fish      0
Name: num_legs, dtype: int64

df['num_specimen_seen'].sample(n=2, random_state=1004)

[Out] 
falcon    10
fish       8
Name: num_specimen_seen, dtype: int64

## Assigning sampling results as Series
samp_Series = df['num_legs'].sample(n=2)
type(samp_Series)

[Out] pandas.core.series.Series

## Assigning sampling results as numpy array
import numpy as np
samp_array = np.array(df['num_legs'].sample(n=2))
type(samp_array)

[Out] numpy.ndarray

samp_array

[Out] array([0, 2])