Generate larger synthetic dataset based on a smaller dataset in Python

问题

I have a dataset with 21000 rows (data samples) and 102 columns (features). I would like to have a larger synthetic dataset generated based on the current dataset, say with 100000 rows, so I can use it for machine learning purposes thereby.

I've been referring to the answer by @Prashant on this post https://stats.stackexchange.com/questions/215938/generate-synthetic-data-to-match-sample-data, but am unable to get it working on generating a larger synthetic dataset for my data.

import numpy as np
from random import randrange, choice
from sklearn.neighbors import NearestNeighbors
import pandas as pd
#referring to https://stats.stackexchange.com/questions/215938/generate-synthetic-data-to-match-sample-data


df = pd.read_pickle('df_saved.pkl')
df = df.iloc[:,:-1] # this gives me df, the final Dataframe which I would like to generate a larger dataset based on. This is the smaller Dataframe with 21000x102 dimensions.


def SMOTE(T, N, k):
# """
# Returns (N/100) * n_minority_samples synthetic minority samples.
#
# Parameters
# ----------
# T : array-like, shape = [n_minority_samples, n_features]
#     Holds the minority samples
# N : percetange of new synthetic samples:
#     n_synthetic_samples = N/100 * n_minority_samples. Can be < 100.
# k : int. Number of nearest neighbours.
#
# Returns
# -------
# S : array, shape = [(N/100) * n_minority_samples, n_features]
# """
    n_minority_samples, n_features = T.shape

    if N < 100:
       #create synthetic samples only for a subset of T.
       #TODO: select random minortiy samples
       N = 100
       pass

    if (N % 100) != 0:
       raise ValueError("N must be < 100 or multiple of 100")

    N = N/100
    n_synthetic_samples = N * n_minority_samples
    n_synthetic_samples = int(n_synthetic_samples)
    n_features = int(n_features)
    S = np.zeros(shape=(n_synthetic_samples, n_features))

    #Learn nearest neighbours
    neigh = NearestNeighbors(n_neighbors = k)
    neigh.fit(T)

    #Calculate synthetic samples
    for i in range(n_minority_samples):
       nn = neigh.kneighbors(T[i], return_distance=False)
       for n in range(N):
          nn_index = choice(nn[0])
          #NOTE: nn includes T[i], we don't want to select it
          while nn_index == i:
             nn_index = choice(nn[0])

          dif = T[nn_index] - T[i]
          gap = np.random.random()
          S[n + i * N, :] = T[i,:] + gap * dif[:]

    return S

df = df.to_numpy()
new_data = SMOTE(df,50,10) # this is where I call the function and expect new_data to be generated with larger number of samples than original df.

The traceback of the error I get is mentioned below:-

Traceback (most recent call last):
  File "MyScript.py", line 66, in <module>
    new_data = SMOTE(df,50,10)
  File "MyScript.py", line 52, in SMOTE
    nn = neigh.kneighbors(T[i], return_distance=False)
  File "/trinity/clustervision/CentOS/7/apps/anaconda/4.3.31/3.6-VE/lib/python3.5/site-packages/sklearn/neighbors/base.py", line 393, in kneighbors
    X = check_array(X, accept_sparse='csr')
  File "/trinity/clustervision/CentOS/7/apps/anaconda/4.3.31/3.6-VE/lib/python3.5/site-packages/sklearn/utils/validation.py", line 547, in check_array
    "if it contains a single sample.".format(array))
ValueError: Expected 2D array, got 1D array instead:

I know that this error (Expected 2D array, got 1D array) is occurring on the line nn = neigh.kneighbors(T[i], return_distance=False). Precisely, when I call the function, T is the numpy array of shape (21000x102), my data which I convert from a Pandas Dataframe to a numpy array. I know that this question may have some similar duplicates, but none of them answer my question. Any help in this regard would be highly appreciated.

回答1:

So what T[i] is giving it is an array with shape (102, ).

What the function expects is an array with shape (1, 102).

You can get this by calling reshape on it:

nn = neigh.kneighbors(T[i].reshape(1, -1), return_distance=False)

In case you're not familiar with np.reshape, The 1 says that the first dimension should be size 1, and the -1 says that the second dimension should be what ever size numpy can broadcast it to; in this case the original 102.

回答2:

May be of use for you

SMOTE and other advanced over_sampling techniques

This package imblearn has sklearn like API and lots of oversampling techniques.

回答3:

I had same problem. I researched for a while but could not find an appropriate solution then I tried to apply my own solution to this problem. It helped me and I hope it will work for all who has same problem.

columns = df.columns.to_numpy()
iteration_count = 30
new_df = pd.DataFrame(columns=columns)

for i in range(iteration_count):
    for k in df.iterrows():
        data_obj = {}
        for j in range(columns.size):
            random_index = np.random.randint(0,13, dtype='int')
            data_obj[columns[j]] = df.loc[random_index][columns[j]]
        new_df = new_df.append(data_obj, ignore_index=True)

df = df.append(new_df, ignore_index=True)

来源：https://stackoverflow.com/questions/55027404/generate-larger-synthetic-dataset-based-on-a-smaller-dataset-in-python

标签

python

machine-learning

scikit-learn

imputation