De-duplicate a bunch of identical code

[jbrockmendel]

The first half of irf_errband_mc is copy/paste identical to irf_resim, so let's go ahead and use that instead.

Also make a helper method for a commonly-used expression in irf.

This does not:

• clean up any formatting
• fix the AttributeError that irf_errband_mc or irf_resim would raise if it were ever called.
• fix the TypeError that irf_errband_mc would raise if it that AttributeError were fixed.

[coveralls]

Coverage increased (+0.03%) to 82.723% when pulling 3c2b110 on jbrockmendel:resim into 98d67db on statsmodels:master.

[codecov-io]

Codecov Report

Merging #4458 into master will increase coverage by 0.02%.
The diff coverage is 45.45%.

@@            Coverage Diff             @@
##           master    #4458      +/-   ##
==========================================
+ Coverage   80.06%   80.09%   +0.02%     
==========================================
  Files         560      560              
  Lines       84369    84347      -22     
  Branches     9559     9554       -5     
==========================================
+ Hits        67551    67554       +3     
+ Misses      14597    14572      -25     
  Partials     2221     2221

Impacted Files	Coverage Δ
statsmodels/tsa/vector_ar/var_model.py	85.11% <0%> (+1.58%)	⬆️
statsmodels/tsa/vector_ar/irf.py	54.36% <50%> (+2.4%)	⬆️
statsmodels/stats/descriptivestats.py	24.13% <0%> (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 98d67db...3c2b110. Read the comment docs.

[jbrockmendel]

@ChadFulton pls take a look when you get a chance

[ChadFulton]

LGTM, I think very straightforward, will merge tonight (or tomorrow if I forget) unless I hear otherwise.

[josef-pkt]

LGTM

[ChadFulton]

Thanks @jbrockmendel for cleanup and @josef-pkt for review!