In [12]:

    

import numpy as np
from validata import check


    

In [13]:

    

np.random.seed(0)

data = np.random.randn(100,4)


    

In [17]:

    

reload(check)
check.check(data)


    

In [18]:

    

data[:,0] = 0
check.check(data)


    

In [ ]:

    

data[:,0] = np.random.randn(100)
for i in range(1,4):
    data[:,i] = data[:,0]+np.random.randn(100)*1e-6
check.check(data)


    

In [53]:

    

labels = np.random.randint(0,2,100)


    

In [56]:

    

from sklearn.preprocessing import OneHotEncoder

onehot_labels = OneHotEncoder(n_values=2).fit_transform(labels[:,None]).toarray()


    

In [68]:

    

assert check.labels_are_one_hot(onehot_labels) == True
assert check.labels_are_one_hot(labels) == False


    

In [70]:

    

check.check_labels_are_1d_or_2d(onehot_labels)
check.check_labels_are_1d_or_2d(labels)


    

In [90]:

    

a = np.array(['a']*30)


    

In [72]:

    

check.check_all_classes_used(onehot_labels)