CODEWITHSUNDEEP
pythoncsvmachine-learningscikit-learndataset
Yuedong HU
Yuedong HU

Reputation: 93

How to create my own datasets using in scikit-learn?

I want to create my own datasets, and use it in scikit-learn. Scikit-learn has some datasets like 'The Boston Housing Dataset' (.csv), user can use it by:

from sklearn import datasets 
boston = datasets.load_boston()

and codes below can get the data and target of this dataset:

X = boston.data
y = boston.target

The question is how to create my own dataset and can be used in that way? Any answers is appreciated, Thanks!

Upvotes: 8

Views: 11470

Answers (2)

Tonechas
Tonechas

Reputation: 13723

Here's a quick and dirty way to achieve what you intend:

my_datasets.py

import numpy as np
import csv
from sklearn.utils import Bunch

def load_my_fancy_dataset():
    with open(r'my_fancy_dataset.csv') as csv_file:
        data_reader = csv.reader(csv_file)
        feature_names = next(data_reader)[:-1]
        data = []
        target = []
        for row in data_reader:
            features = row[:-1]
            label = row[-1]
            data.append([float(num) for num in features])
            target.append(int(label))
        
        data = np.array(data)
        target = np.array(target)
    return Bunch(data=data, target=target, feature_names=feature_names)

my_fancy_dataset.csv

feature_1,feature_2,feature_3,class_label
5.9,1203,0.69,2
7.2,902,0.52,0
6.3,143,0.44,1
-2.6,291,0.15,1
1.8,486,0.37,0

Demo

In [12]: import my_datasets

In [13]: mfd = my_datasets.load_my_fancy_dataset()

In [14]: X = mfd.data

In [15]: y = mfd.target

In [16]: X
Out[16]: 
array([[ 5.900e+00,  1.203e+03,  6.900e-01],
       [ 7.200e+00,  9.020e+02,  5.200e-01],
       [ 6.300e+00,  1.430e+02,  4.400e-01],
       [-2.600e+00,  2.910e+02,  1.500e-01],
       [ 1.800e+00,  4.860e+02,  3.700e-01]])

In [17]: y
Out[17]: array([2, 0, 1, 1, 0])

In [18]: mfd.feature_names
Out[18]: ['feature_1', 'feature_2', 'feature_3']

Upvotes: 5

user10071963
user10071963

Reputation:

Assuming you have your data loaded into memory as a 2D array, there is an easy way to do this with OneHotEncoder:

from sklearn import svm
from sklearn.preprocessing import OneHotEncoder

data = [['Male', 1, True], ['Female', 3, True], ['Female', 2, False], ['Male', 2, False]]
y = [0, 1, 1, 1]  # expected outputs
enc = OneHotEncoder(drop='if_binary')  # create encoder obj that drops unneeded columns on binary inputs
X = enc.fit_transform(data).toarray()  # vectorize input data
clf = svm.SVC(gamma=0.001, C=100.)  # create classifier obj
clf.fit(X[:-1], y[:-1])  # fit model using training data (all but last entry)

out = clf.predict(X[-1:])  # resulting prediction (last entry only)

Upvotes: 0

Related Questions