Default of "allow_downcast" is True, like previous behaviour.
However, default of "allow_input_downcast" is False, so no precision is
silently lost when passing a parameter to a Theano function.

Enable sparse elemwise multiplication by a something that can be converted to a tensor but is not a tensor or a sparse. Test it(and add test for all sparse mul at the same time)

using Variable's comparison operators

Need to swap inheritence order for class ScalarVariable: inherit
_scalar_py_operators first, then Variable

inc_subtensor/set_subtensor.

* changed inheritance order of TensorVariable to include _py_operators before
Variable, so that TensorVariable can use the comparison operators defined in
_py_operators

inc_subtensor/set_subtensor, as the previous implementation is now deprecated.

import the version of theano and nothing else. In some case we could import the version from another project.

the Theano type

refactored TensorSharedVariable test of shared options to be able to reused them for CudaNdarrayShared.

to check for input aliasing and removed the transitivity assumption on the
aliasing.

mutable or borrowed