use distributed processing from DiDa

[exaexa]

After the performance fixes in DistributedData (JuliaLang/julia#44671 (comment), LCSB-BioCore/DistributedData.jl#40 ) it no longer seems vital to maintain a custom clone of the functionality in GigaSOM. This switches to all functions from DistributedData.

In the process of cleaning I also axed visualizations/ as they were not even included in the code (:rofl: ) and bumped to 0.7 so that people may start depending on this.

There might be various leftovers from original gigasom-internal DiDa, esp in docs. I checked a lot with sed but still kinda betting on quality reviews. :D

[laurentheirendt]

Difficult to check every line of it, but looks good as far as I can judge

[laurentheirendt]

Why is CodeCov not responding? Would be interesting to check the testing coverage in this case

[exaexa]

I want Oliver to actually check this in his settings, let's merge after he acks that it works for him.

Why is CodeCov not responding? Would be interesting to check the testing coverage in this case

good question, I've got no idea why it works as it works tbh. Maybe it doesn't care about merging to develop?

..............let's make our own with Julia'n'stuff? :D

[codecov]

Codecov Report

Merging #203 (567ceff) into develop (42af827) will increase coverage by 2.29%.
The diff coverage is 97.95%.

@@             Coverage Diff             @@
##           develop     #203      +/-   ##
===========================================
+ Coverage    91.47%   93.76%   +2.29%     
===========================================
  Files           11        8       -3     
  Lines          598      385     -213     
===========================================
- Hits           547      361     -186     
+ Misses          51       24      -27     

Impacted Files	Coverage Δ
src/base/trainutils.jl	95.12% <0.00%> (+0.25%)	⬆️
src/analysis/core.jl	94.73% <100.00%> (+0.21%)	⬆️
src/analysis/embedding.jl	95.60% <100.00%> (ø)
src/base/dataops.jl	100.00% <100.00%> (+2.10%)	⬆️
src/base/structs.jl	100.00% <100.00%> (ø)
src/io/input.jl	91.13% <100.00%> (+0.11%)	⬆️
src/io/process.jl	87.93% <100.00%> (ø)
src/io/splitting.jl	100.00% <100.00%> (ø)
... and 1 more

---

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 b784d10...567ceff. Read the comment docs.

[laurentheirendt]

Great - thx for fixing Codecov!

[exaexa]

So this actually seems to work just right on ULHPC so I'm going to fastforward a bit :D @oHunewald lmk in case you hit any issues.