Inconsistent presentation of DC in zoltar-truth.csv

[JohnBurant]

Perhaps I'm missing something, but this seems like a bug to me:

(1) DC is present in zoltar-truth.csv as 2-digit FIPS 11 and 5-digit FIPS 11001.
(2) For 11, only 'cum death' shows as target
(3) For both 11 and 11001 both 'inc death' and 'cum death' show as targets
(4) For 11 and 11001 'cum death' targets match
(5) The value for 'inc death' target shown for 11001 is always TWICE the difference between subsequent timezero '1 wk ahead cum death' target values.

Example:

Cumulative deaths:
65143 2020-08-23 11 1 wk ahead cum death 605.0 state
65144 2020-08-23 11001 1 wk ahead cum death 605.0 other
66709 2020-08-30 11 1 wk ahead cum death 611.0 state
66710 2020-08-30 11001 1 wk ahead cum death 611.0 other
69319 2020-09-06 11 1 wk ahead cum death 616.0 state
69320 2020-09-06 11001 1 wk ahead cum death 616.0 other
70189 2020-09-13 11 1 wk ahead cum death 619.0 state
70190 2020-09-13 11001 1 wk ahead cum death 619.0 other
73669 2020-09-20 11 1 wk ahead cum death 624.0 state
73670 2020-09-20 11001 1 wk ahead cum death 624.0 other
140485 2020-09-27 11 1 wk ahead cum death 629.0 state
140486 2020-09-27 11001 1 wk ahead cum death 629.0 other

But incident deaths:
227379 2020-08-30 11001 1 wk ahead inc death 12.0 other
229944 2020-09-06 11001 1 wk ahead inc death 10.0 other
230799 2020-09-13 11001 1 wk ahead inc death 6.0 other
234219 2020-09-20 11001 1 wk ahead inc death 10.0 other
299883 2020-09-27 11001 1 wk ahead inc death 10.0 other

So actually the incremental deaths were 6, 5, 3, 5 not 12, 10, 6, 10. The former show up in the viz.

Ignore leftmost numbers, just my pd.DataFrame index.

[Abdul Hannan]

Hi @JohnBurant , thanks for reporting this! The double truth data values for DC is fixed in #1700. You can run these checks on the new zoltar-truth.csv and confirm.

[Abdul Hannan]

Closing it as it is fixed now.