Rename and reorder difference table columns

[martinholmer]

This pull request tries to make difference table columns more consistently named and in a more sensible column order. These changes were made after dealing with the questions of a TaxBrain user in issue #1581.

[codecov-io]

Codecov Report

Merging #1584 into master will not change coverage.
The diff coverage is 100%.

@@          Coverage Diff           @@
##           master   #1584   +/-   ##
======================================
  Coverage     100%    100%           
======================================
  Files          37      37           
  Lines        2738    2739    +1     
======================================
+ Hits         2738    2739    +1

Impacted Files	Coverage Δ
taxcalc/utils.py	100% <100%> (ø)	⬆️

---

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 30c4567...d0a4f67. Read the comment docs.

[martinholmer]

Analyzing the EITC reform in the BrownKhanna.json file on the master branch, the create_difference_table(..., groupby='weighted_deciles', income_measure='expanded_income', tax_to_diff='combined')` returns a Pandas DataFrame that looks like this:

        tax_cut  tax_inc       count     mean       tot_change perc_inc perc_cut share_of_change perc_aftertax
0     2,207,709        0  16,762,370     -122   -2,037,617,844    0.00%   13.17%           1.49%        -2.80%
1     7,077,443        0  16,763,990   -1,004  -16,838,665,691    0.00%   42.22%          12.34%       -10.28%
2     7,774,019        0  16,762,079   -1,530  -25,647,234,620    0.00%   46.38%          18.80%        -9.23%
3    10,497,689        0  16,763,669   -1,878  -31,475,230,797    0.00%   62.62%          23.07%        -8.18%
4    11,947,402        0  16,763,005   -1,590  -26,651,830,336    0.00%   71.27%          19.53%        -5.28%
5     6,779,702        0  16,763,313   -1,177  -19,728,537,903    0.00%   40.44%          14.46%        -3.00%
6     4,411,075        0  16,763,828     -733  -12,281,009,658    0.00%   26.31%           9.00%        -1.48%
7     1,287,857        0  16,762,557     -106   -1,778,933,626    0.00%    7.68%           1.30%        -0.17%
8            46        0  16,763,592       -0          -11,423    0.00%    0.00%           0.00%        -0.00%
9             0        0  16,763,403        0                0    0.00%    0.00%          -0.00%         0.00%
sums 51,982,943        0 167,631,805      n/a -136,439,071,897      n/a      n/a         100.00%           n/a

The same utility function call on this development branch returns a Pandas DataFrame that looks like this:

           count    tax_cut perc_cut  tax_inc perc_inc     mean       tot_change share_of_change perc_aftertax
0     16,762,370  2,207,709   13.17%        0    0.00%     -122   -2,037,617,844           1.49%        -2.80%
1     16,763,990  7,077,443   42.22%        0    0.00%   -1,004  -16,838,665,691          12.34%       -10.28%
2     16,762,079  7,774,019   46.38%        0    0.00%   -1,530  -25,647,234,620          18.80%        -9.23%
3     16,763,669 10,497,689   62.62%        0    0.00%   -1,878  -31,475,230,797          23.07%        -8.18%
4     16,763,005 11,947,402   71.27%        0    0.00%   -1,590  -26,651,830,336          19.53%        -5.28%
5     16,763,313  6,779,702   40.44%        0    0.00%   -1,177  -19,728,537,903          14.46%        -3.00%
6     16,763,828  4,411,075   26.31%        0    0.00%     -733  -12,281,009,658           9.00%        -1.48%
7     16,762,557  1,287,857    7.68%        0    0.00%     -106   -1,778,933,626           1.30%        -0.17%
8     16,763,592         46    0.00%        0    0.00%       -0          -11,423           0.00%        -0.00%
9     16,763,403          0    0.00%        0    0.00%        0                0          -0.00%         0.00%
sums 167,631,805 51,982,943      n/a        0      n/a      n/a -136,439,071,897         100.00%           n/a

[MattHJensen]

This ordering is much better. Thanks @martinholmer.