Adding back the list...
It works for me:
In [15]: from sklearn.preprocessing import OneHotEncoder
In [16]: enc = OneHotEncoder()
In [17]: enc.fit([[0, 0, 3], [1, 1, 0], [0, 2, 1], [1, 0, 2]])
Out[17]:
OneHotEncoder(categorical_features='all', dtype=<type 'float'>,
handle_unknown='error', n_values='auto', sparse=True)
In [18]: enc.transform([[0, 1, 1]]).toarray()
Out[18]: array([[ 1., 0., 0., 1., 0., 0., 1., 0., 0.]])
In [19]: import cPickle as pickle
In [20]: s = pickle.dumps(enc)
In [21]: enc_deser = pickle.loads(s)
In [22]: enc_deser.transform([[0, 1, 1]]).toarray()
Out[22]: array([[ 1., 0., 0., 1., 0., 0., 1., 0., 0.]])
Perhaps provide the stack trace and example code to reproduce if possible?
On Wed, Nov 18, 2015 at 9:53 AM, Startup Hire <[email protected]>
wrote:
> Hi Nick,
>
> Quick doubt:
>
> I saved the OneHotEncoder "fit" function ie in
> b=enc.fit(test) , I saved b
>
> Then, I used b.transform(newdata) , but the error
> now comes because of # of rows in test and newdata being different
>
> Am I doing something wrong in saving the OneHotEncoder and reusing it?
>
> Regards,
> Sanant
>
> On Wed, Nov 18, 2015 at 11:24 AM, Nick Pentreath <[email protected]
> > wrote:
>
>> No problem - hope you solve your issue :)
>>
>> —
>> Sent from Mailbox <https://www.dropbox.com/mailbox>
>>
>>
>> On Wed, Nov 18, 2015 at 5:47 AM, Startup Hire <[email protected]>
>> wrote:
>>
>>> Thanks for your explanation!
>>>
>>>
>>
>
------------------------------------------------------------------------------
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
