search

QuantumSavory / QuantumSavory.jl

A full stack simulator of quantum hardware, from the low-level analog physics to high-level network dynamics. Includes discrete event simulator, symbolic representation for quantum object, and works with many backend simulators.
https://quantumsavory.github.io/QuantumSavory.jl/
MIT License
30 stars 11 forks source link

update lower bounds of dependencies and enable downgrade CI job #108

Closed ba2tro closed 6 months ago

codecov[bot] commented 6 months ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 71.11%. Comparing base (1a0b40d) to head (f217fa5).

Additional details and impacted files ```diff @@ Coverage Diff @@ ## master #108 +/- ## ======================================= Coverage 71.11% 71.11% ======================================= Files 37 37 Lines 1475 1475 ======================================= Hits 1049 1049 Misses 426 426 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

ba2tro commented 6 months ago

This shouldn't happen: https://github.com/QuantumSavory/QuantumSavory.jl/actions/runs/8008861646/job/21876235207?pr=108#step:6:60 Since both Clifford and Savory have a lower bound set to 0.3.3 for QuantumInterface

ba2tro commented 6 months ago

Looks like its checking v0.8.0 for QuantumClifford which has lower bound set to 0.1.0 for QuantumInterface, https://github.com/QuantumSavory/QuantumClifford.jl/blob/7da8d351320b73e3f9461fa68b65c60ec03da3ea/Project.toml#L46

so we can either lower the bound for QuantumInterface or increase the bound for QuantumClifford which is currently set to 0.8 and is currently at 0.8.19

I think there could be a lot of such adjustments to be made here for other dependencies, to make the downgrade ci pass, but it will be good to set realistic lower bounds

Krastanov commented 6 months ago

I would suggest always going for increasing bounds, in this case increasing the one for QuantumClifford.

Krastanov commented 6 months ago

@Abhishek-1Bhatt , the change to Distributions's lower bound is a bit awkward -- I made it because 0.25.0 does not support newer versions of FillArrays, which I learnt by using github blame on their project.toml file. This is probably not the most elegant and efficient way to run the downstream CI job, but it is better than nothing and it is not a high priority to make it better.