-
When operating on arrays with singleton dimensions, numpy.einsum does not always enforce shape matching over summed dimensions. Instead it can return a surprising result. Minimal example:
```
import…
-
a = np.ones(shape=(5,1))
b = np.array([1,2])
np.einsum('ij,j->', a, b, optimize=False)
# 15.0
np.einsum('ij,j->', a, b, optimize=True)
# ValueError("Size of label '%s' for…
-
Hello,
I would like to apply SCORE to some very large datasets (e.g. DAGs with d=approx 200 nodes and n=approx 20000 samples).\
I am currently blocked by a lack of memory error inside the procedur…
-
In `Einsum.jl`
```julia
@einsum n[i,j,k] = hg[i,m]*hg[j,m]*hg[k,m]
```
defines a star shaped contraction, which will produce a rank 3 tensor.
However, `@tensor` in `TensorOperations.jl` will ra…
-
I need help to use Quarkus.
-
When trying to sum over axes using the ellipsis notation, numpy throws an error.
Minimum code for replication:
```
>>> import numpy as np
>>> a = np.random.rand(2, 3, 5)
>>> np.einsum("...i->i"…
-
While working in #3, I just noticed that panels are not being generated properly since a local chord shift is being introduced:
https://github.com/partmor/ezaero/blob/a7c1f31b191ea605e9a9802c645a0f…
-
Is there an example of having arraymancer look at a folder containing multiple classes in their own folders, specifying a train / test split, and generating a model based on that? Finally saving the m…
-
A list of PyTorch 1.7 features.
Items are checked if we have something more or less equivalent in Flux or in the julia ecosystem and supported by Flux.
This list is not complete, it comes from a r…
-
I hope to track and change summation order in `@tullio`. What is the default order when tullio meet repeated indices? And is it possible to change them?