Features/torch proxy

[ClaudiaComito]

Description

I propose to implement a DNDarray method to return a 1-byte torch.Tensor strided as the original DNDarray global shape. At @ben-bou 's suggestion, we have been using this "proxy" tensor for sanitation, as it allows us to easily bypass many of the instance checks. It is also quite practical to assess the shape of the output DNDarray of any operation. See getitem/setitem, or manipulations.tile() for examples.

Issue/s resolved: None. Implemented with an eye on long-term maintenance / minimizing repetitions / simplifying code.

Changes proposed:

• Implement DNDarray.__torch_proxy__()
• Implement tests
• replace "proxy" definitions in dndarray and manipulations with DNDarray.__torch_proxy__() calls
• enforce DNDarray.size returning int even if DNDarray.ndim is 0

Type of change

• New feature (non-breaking change which adds functionality)

Due Diligence

• All split configurations tested
• Multiple dtypes tested in relevant functions
• Documentation updated (if needed)
• Updated changelog.md under the title "Pending Additions"

Does this change modify the behaviour of other functions? If so, which?

no

[codecov]

Codecov Report

Merging #856 (be8979d) into master (e2f75c3) will increase coverage by 0.05%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #856      +/-   ##
==========================================
+ Coverage   95.40%   95.45%   +0.05%     
==========================================
  Files          64       64              
  Lines        9380     9382       +2     
==========================================
+ Hits         8949     8956       +7     
+ Misses        431      426       -5     

Flag	Coverage Δ
gpu	94.62% <100.00%> (+0.05%)	⬆️
unit	91.22% <100.00%> (+0.05%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
heat/core/dndarray.py	97.31% <100.00%> (+<0.01%)	⬆️
heat/core/manipulations.py	98.94% <100.00%> (+0.37%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e2f75c3...be8979d. Read the comment docs.

[ClaudiaComito]

rerun tests