Use proper sample IDs inside feature tables

[xtrojak]

All feature tables contained arbitrary (resp. based on input order) integer sample IDs. The original name of the sample was lost during the feature extraction process. This PR reintroduces them again and uses them to identify the sample across the computation.

Requires files update here.

[maximskorik]

@xtrojak, I've checked out to this branch and fetched the new test data but the following tests fail: adjust_time, compute_clusters, feature-align, hybrid, recover-weaker, unsupervised. Some of them due to actual output not being equal to expected and some (recover-weaker and adjust-time) with this error message:

...
Can't combine `..1$sample_id` <double> and `..2$sample_id` <character>.

or

...
Can't combine `..1$sample_id` <character> and `..2$sample_id` <integer>.

Do you have any idea what I might be missing? I've run the tests in the devcontainer.

[maximskorik]

The same tests fail with devtools::check('.', error_on = 'error'):

[ FAIL 14 | WARN 21 | SKIP 4 | PASS 24 ]
Error: Test failures
In addition: Warning messages:
1: In features$cluster : closing unused connection 6 (<-localhost:11634)
2: In features$cluster : closing unused connection 5 (<-localhost:11634)
Execution halted

1 error ✖ | 6 warnings ✖ | 7 notes ✖
Error: R CMD check found ERRORs
Execution halted

[xtrojak]

@maximskorik Are you sure you fetched the new test data properly? If you just run the wget commands, the files are not overwritten but a new copy is placed alongside them (with a .1, .2 ... suffix).

I suppose we should add -O parameter to the readme.

[maximskorik]

@maximskorik Are you sure you fetched the new test data properly? If you just run the wget commands, the files are not overwritten but a new copy is placed alongside them (with a .1, .2 ... suffix).

I suppose we should add -O parameter to the readme.

Thanks for the advice! That must've been my problem.

[maximskorik]

All tests pass and the new test-data looks good. Thanks, @xtrojak!
This can be merged. Also if you could update the call to wget with -O in README.md that would be nice.