How to add a dummy dimension?

[madanh]

To use Ordered() I have to have my vector of dimension N, but then I do some manipulation with the result and the rest of the function is written to accept (N,1) shaped stuff (subclasses of theano tensors, I suppose). What is the proper way to add a dummy dimension?

[junpenglao]

There is a discussion here

Order statistics in PyMC3

OK, I cooked up something that kind of work: import numpy as np import pymc3 as pm import theano.tensor as tt # data K = 5 # number of items being ranked J = 20 # number of raters yreal = np.argsort(np.random.randn(K, 1), axis=0) print(yreal) y = np.argsort(yreal + np.random.randn(K, J), axis=0) print(y) class Ordered2D(pm.distributions.transforms.ElemwiseTransform): name = "ordered" def forward(self, x): out = tt.zeros(x.shape) out = tt.inc_subtensor(out[0,:], x[0,:]…