Improve handling of the index columns

[olejandro]

This PR removes generation of all know index columns; selective generation is done instead. Columns without any values are dropped. This allows cutting down on the processing time.

[github-actions]

Regression test results on commit d1e2776

       Benchmark    Time (s)         GDX Diff     Accuracy       Correct    Additional
----------------  ----------  ---------------  -----------  ------------  ------------
   DemoS_001-all   3.2   3.6    OK         OK  100.0 100.0    118    118      3      3
   DemoS_002-all   3.7   4.3    OK         OK  100.0 100.0    344    344      3      3
   DemoS_003-all   4.2   4.3    OK         OK  100.0 100.0    633    633      6      6
       DemoS_004   4.3   4.4    OK         OK  100.0 100.0    662    662     12     12
      DemoS_004a   4.5   5.2    OK         OK  100.0 100.0    665    665     12     12
      DemoS_004b   4.4   4.5    OK         OK  100.0 100.0    665    665     12     12
   DemoS_004-all   4.6   5.0    OK         OK  100.0 100.0    667    667     12     12
   DemoS_005-all   5.1   4.9    13         13   99.4  99.4   1160   1160     12     12
   DemoS_006-all   6.2   6.1    13         13   99.4  99.4   1258   1258     12     12
   DemoS_007-all   7.2   8.1    13         13   99.6  99.6   2155   2155     12     12
DemoS_007-all-1r   5.8   5.6    14         14   99.5  99.5   1179   1179     12     12
   DemoS_008-all   9.8   9.1    13         13   99.7  99.7   5333   5333     18     18
   DemoS_009-all  10.9   9.5    26         26   99.7  99.7   5807   5807     29     29
   DemoS_010-all  11.3   9.2   822        822   88.2  88.2   6139   6139     29     29
   DemoS_011-all  11.3   9.3   854        854   87.9  87.9   6150   6150     29     29
   DemoS_012-all  11.9   9.8   901        901   88.0  88.0   6306   6306     53     53
    TIMES-IE-all  33.5  26.9  3211       3131   97.5  97.5  42384  42460   2826   2826
    TIMES-IE-NoM  30.2  24.1   363        363   99.3  99.3  40528  40528    632    632
    TIMES-IE-MCB  32.2  25.4  1186       1106   97.6  97.6  41991  42067    632    632
    TIMES-NZ-KEA  60.0  33.6  4004       4004   98.2  98.2  76362  76362    472    472
    TIMES-NZ-TUI  48.3  23.3  4038       4038   98.2  98.2  75988  75988    488    488
 (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:373")
2025-01-23 02:00:29.103 | INFO     : Total runtime: 235.94s (main: 312.64s) (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:397")
2025-01-23 02:00:29.103 | INFO     : Change in runtime (negative == faster): -76.69s (-24.5%) (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:398")
2025-01-23 02:00:29.103 | INFO     : Change in correct rows (higher == better): +152 (+0.0%) (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:405")
2025-01-23 02:00:29.103 | INFO     : Change in additional rows: +0 (+0.0%) (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:412")
2025-01-23 02:00:29.104 | SUCCESS  : No regressions. You're awesome! (__main__:MainThread:pid-5898 "/home/runner/work/xl2times/xl2times/xl2times/utils/run_benchmarks.py:426")

This comment will be updated when new commits are added to the PR.