ChaLearn Automatic Machine Learning Challenge (AutoML) Forum

Go back to competition Back to thread list Post in this thread

> cadata fails in AutoML Round 0 ?

I have an issue on Round 0 submission :
=========== Cadata Training cycle 0 ================
[+] Number of estimators: 1
[+] Fitting success, time spent so far 0.01 sec
[+] Size of trained model 72.00 bytes
Traceback (most recent call last):
File "", line 320, in <module>
Y_test = M.predict(['X_test'])
File "/data/AutoML-data/newRound/sample_code_submission/lib/", line 104, in predict
prediction = np.vstack((prediction,self.predict_method(X[c,:])))
File "/home/bruno/anaconda/lib/python2.7/site-packages/numpy/core/", line 230, in vstack
return _nx.concatenate([atleast_2d(_m) for _m in tup], 0)
ValueError: all the input array dimensions except for the concatenation axis must match exactly

Any idea ?


Posted by: brunoseznec @ Dec. 17, 2015, 3:33 p.m.

Sorry for late response, all other conversations are on google groups.
I hope you have figured out what is wrong (cadata output is one dimensional and you could use hstack instead). If this is reference submission, it could be problem for AutoML phase 4, when regression will be introduced, and you could notice organizers if that should be corrected.


Posted by: djajetic @ Jan. 14, 2016, 9:35 a.m.
Post in this thread