Synthetic data with SDV and CopulaGAN

data.head()

	Pclass	Sex	Age	SibSp	Fare
0	3	1	22.0	1	7.2500
1	1	0	38.0	1	71.2833
2	3	0	26.0	0	7.9250
3	1	0	35.0	1	53.1000
4	3	1	35.0	0	8.0500

##    Pclass  Sex   Age  SibSp  Parch     Fare
## 0       3    1  22.0      1      0   7.2500
## 1       1    0  38.0      1      0  71.2833
## 2       3    0  26.0      0      0   7.9250
## 3       1    0  35.0      1      0  53.1000
## 4       3    1  35.0      0      0   8.0500

data.describe(include='all')

	Pclass	Sex	Age	SibSp	Parch	Fare
count	891.000000	891.000000	891.000000	891.000000	891.000000	891.000000
mean	2.308642	0.647587	29.758889	0.523008	0.381594	32.204208
std	0.836071	0.477990	13.002570	1.102743	0.806057	49.693429
min	1.000000	0.000000	0.420000	0.000000	0.000000	0.000000
25%	2.000000	0.000000	22.000000	0.000000	0.000000	7.910400
50%	3.000000	1.000000	30.000000	0.000000	0.000000	14.454200
75%	3.000000	1.000000	35.000000	1.000000	0.000000	31.000000
max	3.000000	1.000000	80.000000	8.000000	6.000000	512.329200

##            Pclass         Sex         Age       SibSp       Parch        Fare
## count  891.000000  891.000000  891.000000  891.000000  891.000000  891.000000
## mean     2.308642    0.647587   29.758889    0.523008    0.381594   32.204208
## std      0.836071    0.477990   13.002570    1.102743    0.806057   49.693429
## min      1.000000    0.000000    0.420000    0.000000    0.000000    0.000000
## 25%      2.000000    0.000000   22.000000    0.000000    0.000000    7.910400
## 50%      3.000000    1.000000   30.000000    0.000000    0.000000   14.454200
## 75%      3.000000    1.000000   35.000000    1.000000    0.000000   31.000000
## max      3.000000    1.000000   80.000000    8.000000    6.000000  512.329200

from sdv.tabular import CopulaGAN

model = CopulaGAN()

model.fit(data)

new_data = model.sample(200)

new_data.head()

	Pclass	Sex	Age	SibSp	Parch	Fare
0	1	0	23.113076	1	0	29.937752
1	3	1	29.918377	1	0	5.958775
2	1	1	76.710166	0	0	202.889201
3	3	0	46.581959	1	1	89.833194
4	1	0	33.135297	1	2	118.551008

##    Pclass  Sex        Age  SibSp  Parch        Fare
## 0       1    0  11.797730      3      2   15.963036
## 1       2    0  28.135436      2      2   12.701402
## 2       3    0  26.367627      0      1   10.742400
## 3       3    1  24.116738      0      1  144.611161
## 4       2    0  31.368429      3      1   13.659402

new_data.describe(include='all')

	Pclass	Sex	Age	SibSp	Parch	Fare
count	200.000000	200.000000	200.000000	200.00000	200.000000	200.000000
mean	1.960000	0.540000	37.131076	1.10500	0.975000	69.046022
std	0.831781	0.499648	17.131095	1.03408	1.196467	70.354667
min	1.000000	0.000000	3.409883	0.00000	0.000000	3.135404
25%	1.000000	0.000000	27.213419	0.00000	0.000000	10.772323
50%	2.000000	1.000000	31.788050	1.00000	0.000000	42.555842
75%	3.000000	1.000000	49.398770	2.00000	2.000000	105.964913
max	3.000000	1.000000	78.841709	5.00000	5.000000	294.906651

##            Pclass         Sex         Age       SibSp       Parch        Fare
## count  200.000000  200.000000  200.000000  200.000000  200.000000  200.000000
## mean     1.990000    0.440000   31.717758    0.945000    1.240000   76.998168
## std      0.795654    0.497633   14.745161    0.925341    1.487773   82.340578
## min      1.000000    0.000000   -0.299758    0.000000    0.000000    4.534767
## 25%      1.000000    0.000000   23.051398    0.000000    0.000000   14.265699
## 50%      2.000000    0.000000   30.367405    1.000000    1.000000   36.297745
## 75%      3.000000    1.000000   40.796266    1.000000    2.000000  115.727641
## max      3.000000    1.000000   77.002768    5.000000    6.000000  350.698375

from sdv.evaluation import evaluate

evaluate(new_data, data)

0.2793053940561327

## 0.4487854281948458

model = CopulaGAN(
    field_transformers={
        'Pclass': 'categorical',
        'Sex': 'categorical',
        'Age': 'float',
        'SibSp': 'boolean',
        'Parch': 'integer',
        'Fare': 'float'
    },
    field_distributions={
        'Fare': 'truncated_gaussian'
    }
)

model.fit(data)

new_data = model.sample(200)

new_data.head()

	Pclass	Sex	Age	SibSp	Parch	Fare
0	2	1	55.113425	1	0	-9.755683e-08
1	2	1	42.257149	1	2	7.252636e+00
2	3	1	30.128989	1	0	7.708430e+00
3	2	1	30.838961	1	0	-1.184872e-07
4	3	1	30.370534	1	0	3.530599e+00

new_data.describe(include='all')

	Pclass	Sex	Age	SibSp	Parch	Fare
count	200.000000	200.000000	200.000000	200.000000	200.000000	2.000000e+02
mean	2.265000	0.630000	33.075478	0.715000	0.915000	2.587770e+01
std	0.726231	0.484016	13.472204	0.452547	1.189474	3.599977e+01
min	1.000000	0.000000	-1.122344	0.000000	0.000000	-1.191555e-07
25%	2.000000	0.000000	26.064943	0.000000	0.000000	6.589631e+00
50%	2.000000	1.000000	30.706679	1.000000	0.000000	1.182216e+01
75%	3.000000	1.000000	41.977015	1.000000	2.000000	2.797455e+01
max	3.000000	1.000000	74.046067	1.000000	5.000000	2.014112e+02

evaluate(new_data, data)

0.3533601962024023

Synthetic data with SDV and CopulaGAN

Contents

Backlinks