Don’t define new Ops unless you have to¶
It is usually not useful to define Ops that can be easily implemented using other already existing Ops. For example, instead of writing a “sum_square_difference” Op, you should probably just write a simple function:
from theano import tensor as tt def sum_square_difference(a, b): return tt.sum((a - b)**2)
Even without taking Theano’s optimizations into account, it is likely to work just as well as a custom implementation. It also supports all data types, tensors of all dimensions as well as broadcasting, whereas a custom implementation would probably only bother to support contiguous vectors/matrices of doubles…
Use Theano’s high order Ops when applicable¶
Theano provides some generic Op classes which allow you to generate a lot of Ops at a lesser effort. For instance, Elemwise can be used to make elementwise operations easily whereas DimShuffle can be used to make transpose-like transformations. These higher order Ops are mostly Tensor-related, as this is Theano’s specialty.
Use this list to make sure you haven’t forgotten anything when defining a new Op. It might not be exhaustive but it covers a lot of common mistakes.