SVI notebook errors in Python 3.5

Hi! I’m trying to run the SVI notebook (http://nbviewer.jupyter.org/github/SheffieldML/notebook/blob/master/GPy/SVI.ipynb) on Python 3.5 using the devel branch of GPy. I get the following error at In [4]:

Z = np.random.rand(20,1)

batchsize = 10
m = GPy.core.SVGP(X, Y, Z, GPy.kern.RBF(1) + GPy.kern.White(1), GPy.likelihoods.Gaussian(), batchsize=batchsize)
m.kern.white.variance = 1e-5
m.kern.white.fix()

---------------------------------------------------------------------------

TypeError                                 Traceback (most recent call last)

<ipython-input-4-382f732b700d> in <module>()
      2 
      3 batchsize = 10
----> 4 m = GPy.core.SVGP(X, Y, Z, GPy.kern.RBF(1) + GPy.kern.White(1), GPy.likelihoods.Gaussian(), batchsize=batchsize)
      5 m.kern.white.variance = 1e-5
      6 m.kern.white.fix()


/.../lib/python3.5/site-packages/paramz/parameterized.py in __call__(self, *args, **kw)
     46         self._in_init_ = True
     47         #import ipdb;ipdb.set_trace()
---> 48         self = super(ParametersChangedMeta, self).__call__(*args, **kw)
     49         #logger.debug("finished init")
     50         self._in_init_ = False


/.../lib/python3.5/site-packages/GPy/core/svgp.py in __init__(self, X, Y, Z, kernel, likelihood, mean_function, name, Y_metadata, batchsize, num_latent_functions)
     34             #Make a climin slicer to make drawing minibatches much quicker
     35             self.slicer = climin.util.draw_mini_slices(self.X_all.shape[0], self.batchsize)
---> 36             X_batch, Y_batch = self.new_batch()
     37 
     38         #create the SVI inference method


/.../lib/python3.5/site-packages/GPy/core/svgp.py in new_batch(self)
     90         Return a new batch of X and Y by taking a chunk of data from the complete X and Y
     91         """
---> 92         i = next(self.slicer)
     93         return self.X_all[i], self.Y_all[i]
     94 


/.../lib/python3.5/site-packages/climin/util.py in draw_mini_slices(n_samples, batch_size, with_replacement)
     90     else:
     91         while True:
---> 92             random.shuffle(idxs)
     93             for i in idxs:
     94                 yield slices[i]


/usr/lib/python3.5/random.py in shuffle(self, x, random)
    270                 # pick an element in x[:i+1] with which to exchange x[i]
    271                 j = randbelow(i+1)
--> 272                 x[i], x[j] = x[j], x[i]
    273         else:
    274             _int = int


TypeError: 'range' object does not support item assignment

[wilocw]

This seems to work/not work depending on the version of climin. The PyPI version of climin is different to the Github version. Installing via pip install climin will install climin 0.1a1 from PyPI, but following the instructions on the readthedocs, as follows, will install climin pre-0.1, which should work with Python 3.5+:

$ git clone https://github.com/BRML/climin.git /path/to/climin
$ cd /path/to/climin
$ pip install -e .

[mohammadalimehdizadeh]

i have the same problem with google colab python version3.6 and @wilocw solution work well.