fix: time comparison can't guarantee the accuracy

[zhaoyongjie]

SUMMARY

The time comparison in AA can't guarantee accuracy when applied to multiple series line charts.

Due to the errors of the original algorithm, only consider using the time column to join. Now the join algorithm is improved.

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

After

Before

Time.comparison-before.mov

TESTING INSTRUCTIONS

1. select birth_names dataset
2. switch to time-series chart
3. set time grain to year
4. set time range to 1970 - 2000
5. set metrics to sum__num
6. set group by to state
7. set series limit to 2 for convenience
8. set time shift to 3 years later
9. set calculation type to actual values
10. use the south panel to observe the accuracy of the data

ADDITIONAL INFORMATION

• Has associated issue:
• Required feature flags:
• Changes UI
• Includes DB Migration (follow approval process in SIP-59)
  • Migration is atomic, supports rollback & is backwards-compatible
  • Confirm DB migration upgrade and downgrade tested
  • Runtime estimates and downtime expectations provided
• Introduces new feature or API
• Removes existing feature or API

[codecov]

Codecov Report

Merging #16895 (8e12718) into master (3d8cc15) will decrease coverage by 0.11%.
The diff coverage is 84.15%.

❗ Current head 8e12718 differs from pull request most recent head fc5b43f. Consider uploading reports for the commit fc5b43f to get more accurate results

@@            Coverage Diff             @@
##           master   #16895      +/-   ##
==========================================
- Coverage   77.02%   76.90%   -0.12%     
==========================================
  Files        1021     1021              
  Lines       54754    54849      +95     
  Branches     7470     7481      +11     
==========================================
+ Hits        42173    42183      +10     
- Misses      12335    12419      +84     
- Partials      246      247       +1     

Flag	Coverage Δ
hive	?
mysql	81.85% <89.04%> (+<0.01%)	⬆️
postgres	81.91% <89.11%> (+<0.01%)	⬆️
presto	81.79% <88.43%> (+<0.01%)	⬆️
python	82.17% <89.11%> (-0.24%)	⬇️
sqlite	81.52% <88.35%> (+<0.01%)	⬆️

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

Impacted Files	Coverage Δ
superset-frontend/src/views/CRUD/hooks.ts	47.23% <ø> (ø)
superset/datasets/commands/update.py	94.11% <ø> (ø)
superset/datasets/dao.py	96.50% <ø> (-0.12%)	⬇️
...erset-frontend/src/SqlLab/components/ResultSet.tsx	61.94% <20.00%> (-1.42%)	⬇️
...eFilters/FiltersConfigModal/FiltersConfigModal.tsx	89.42% <64.28%> (-2.80%)	⬇️
...mponents/nativeFilters/FiltersConfigModal/utils.ts	72.22% <66.66%> (ø)
superset/utils/sqllab_execution_context.py	90.55% <66.66%> (-1.83%)	⬇️
superset/sqllab/exceptions.py	80.95% <80.95%> (ø)
superset/sqllab/command.py	82.98% <88.63%> (+0.54%)	⬆️
...onfigModal/FiltersConfigForm/FiltersConfigForm.tsx	73.83% <92.85%> (-0.05%)	⬇️
... and 12 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 3d8cc15...fc5b43f. Read the comment docs.

[villebro]

LGTM and tested to work as expected. Thanks for fixing this and adding comprehensive tests!