Using Area Plots - Specify area to shade and calculate area under curve

[Leonard Scott]

Hello:
I am trying to create twoway Area Plots and I would like to:

1. Specify that only the area below a certain threshold (the set base in the command) is shaded, rather than what seems to be the Stata default to shade areas above and below. Say base is 81. I would only want to shade the areas of dt where level<81
This is the command I used:

twoway (area level dt, base(81))

2. I would then like to calculate an area under the curve for this shaded region. I was considering using the pharmacokinetics -pk- commands, as those seem to be in a similar process. But I didn't know if that is the best method or if there are more appropriate methods.

My data is below (please note just to explain that I have this set as time series and I used tsfill to fill in with missings, which is why there are so many - however I don't think that should affect our methods).

Thank you for any insight!

* Example generated by -dataex-. For more info, type help dataex
clear
input double dt int level
1.9742361e+12 75
1974236160000  .
1974236220000  .
1974236280000  .
1974236340000 72
1.9742364e+12  .
1974236460000  .
1974236520000  .
1974236580000  .
1974236640000 68
1.9742367e+12  .
1974236760000  .
1974236820000  .
1974236880000  .
1974236940000 74
 1.974237e+12  .
1974237060000  .
1974237120000  .
1974237180000  .
1974237240000 74
1.9742373e+12  .
1974237360000  .
1974237420000  .
1974237480000  .
1974237540000 70
1.9742376e+12  .
1974237660000  .
1974237720000  .
1974237780000  .
1974237840000 78
1.9742379e+12  .
1974237960000  .
1974238020000  .
1974238080000  .
1974238140000 82
1.9742382e+12  .
1974238260000  .
1974238320000  .
1974238380000  .
1974238440000 87
1.9742385e+12  .
1974238560000  .
1974238620000  .
1974238680000  .
1974238740000 90
1.9742388e+12  .
1974238860000  .
1974238920000  .
1974238980000  .
1974239040000 96
1.9742391e+12  .
1974239160000  .
1974239220000  .
1974239280000  .
1974239340000 96
1.9742394e+12  .
1974239460000  .
1974239520000  .
1974239580000  .
1974239640000 95
1.9742397e+12  .
1974239760000  .
1974239820000  .
1974239880000  .
1974239940000 93
  1.97424e+12  .
1974240060000  .
1974240120000  .
1974240180000  .
1974240240000 94
1.9742403e+12  .
1974240360000  .
1974240420000  .
1974240480000  .
1974240540000 96
1.9742406e+12  .
1974240660000  .
1974240720000  .
1974240780000  .
1974240840000 90
1.9742409e+12  .
1974240960000  .
1974241020000  .
1974241080000  .
1974241140000 90
1.9742412e+12  .
1974241260000  .
1974241320000  .
1974241380000  .
1974241440000 86
1.9742415e+12  .
1974241560000  .
1974241620000  .
1974241680000  .
1974241740000 85
1.9742418e+12  .
1974241860000  .
1974241920000  .
1974241980000  .
1974242040000  .
end
format %tc dt

[Andrew Musau]

Interpolate (extrapolate) the points in between (at the end) and then have two twoway calls.

* Example generated by -dataex-. To install: ssc install dataex
clear
input double dt int level
1.9742361e+12 75
1974236160000  .
1974236220000  .
1974236280000  .
1974236340000 72
1.9742364e+12  .
1974236460000  .
1974236520000  .
1974236580000  .
1974236640000 68
1.9742367e+12  .
1974236760000  .
1974236820000  .
1974236880000  .
1974236940000 74
 1.974237e+12  .
1974237060000  .
1974237120000  .
1974237180000  .
1974237240000 74
1.9742373e+12  .
1974237360000  .
1974237420000  .
1974237480000  .
1974237540000 70
1.9742376e+12  .
1974237660000  .
1974237720000  .
1974237780000  .
1974237840000 78
1.9742379e+12  .
1974237960000  .
1974238020000  .
1974238080000  .
1974238140000 82
1.9742382e+12  .
1974238260000  .
1974238320000  .
1974238380000  .
1974238440000 87
1.9742385e+12  .
1974238560000  .
1974238620000  .
1974238680000  .
1974238740000 90
1.9742388e+12  .
1974238860000  .
1974238920000  .
1974238980000  .
1974239040000 96
1.9742391e+12  .
1974239160000  .
1974239220000  .
1974239280000  .
1974239340000 96
1.9742394e+12  .
1974239460000  .
1974239520000  .
1974239580000  .
1974239640000 95
1.9742397e+12  .
1974239760000  .
1974239820000  .
1974239880000  .
1974239940000 93
  1.97424e+12  .
1974240060000  .
1974240120000  .
1974240180000  .
1974240240000 94
1.9742403e+12  .
1974240360000  .
1974240420000  .
1974240480000  .
1974240540000 96
1.9742406e+12  .
1974240660000  .
1974240720000  .
1974240780000  .
1974240840000 90
1.9742409e+12  .
1974240960000  .
1974241020000  .
1974241080000  .
1974241140000 90
1.9742412e+12  .
1974241260000  .
1974241320000  .
1974241380000  .
1974241440000 86
1.9742415e+12  .
1974241560000  .
1974241620000  .
1974241680000  .
1974241740000 85
1.9742418e+12  .
1974241860000  .
1974241920000  .
1974241980000  .
1974242040000  .
end
format %tc dt

ipolate level dt, g(level2) epolate
set scheme s1mono
twoway (area level2 dt, base(81) lc(black) fc(gray)) ///
(area level2 dt if level2>81, base(81) lc(black) fc(white)), ///
xtitle("") xlab(, labsize(small)) ytitle("Some description") ///
xsc(range(. `=tc(24jul2022 00:40)')) leg(off)

Res.:

[Andrew Musau]

#2: For calculating area, see

help integ

Time is measured in milliseconds and this is multiplied by whatever units you have for your y variable. So you need to think about any rescaling that you need to do. If you want to report the area underneath the threshold as a proportion of the entire area, then that is fine as it is unitless.

[Leonard Scott]

Thank you Andrew Musau this was very helpful.