Bulk commit to DB and Cache IDs

[simonhkswan]

Inserting data into the database was quite slow, especially when we were commiting each individual
value into the database in a diff-correlation metric. For every value we were also querying the
database for a dataset_id and a metric_id.

So to speed up the process we now use a single session.commit() for each of the DataFrame metrics.
We also use cachetools on the database/metric id functions. It's much faster locally and likely
and even better speedup when the database is remote.

[codecov]

Codecov Report

Attention: 7 lines in your changes are missing coverage. Please review.

Comparison is base (68fbadb) 79.9% compared to head (edb0d2e) 79.0%.

❗ Current head edb0d2e differs from pull request most recent head 0be0676. Consider uploading reports for the commit 0be0676 to get more accurate results

Additional details and impacted files

@@           Coverage Diff            @@
##           master    #162     +/-   ##
========================================
- Coverage    79.9%   79.0%   -0.9%     
========================================
  Files          11      11             
  Lines         926     956     +30     
  Branches      145     155     +10     
========================================
+ Hits          740     756     +16     
- Misses        142     149      +7     
- Partials       44      51      +7     

Files	Coverage Δ
src/insight/metrics/metrics_usage.py	94.1% <100.0%> (ø)
src/insight/metrics/base.py	72.1% <93.7%> (-2.7%)	⬇️
src/insight/database/utils.py	73.1% <71.4%> (-3.2%)	⬇️

[sonarqubecloud]

Quality Gate passed

Kudos, no new issues were introduced!

0 New issues
0 Security Hotspots
75.3% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud