Is it possible to make plots optional for scales?
I am currently benchmarking multiple data structures (compared to List and Array) at exponential size scales (1 to 100K) on different operations. I have one file per benchmarked operation. The plots are cluttering a bit the nice quick overview of all results that used to fit roughly in the page (for exploration I only use 3 scales, 10 to 1000).
The code looks like this:
main : BenchmarkProgram
main =
program <|
describe "Append" <|
[ lists
, hamtArrays
, uint8Arrays
, float64Arrays
]
lists : Benchmark
lists =
Constants.sizeScales
|> List.map (\size -> ( size, List.repeat size 0 ))
|> List.map (\( size, list ) -> ( toString size, \_ -> List.append list list ))
|> scale "List"
hamtArrays : Benchmark
hamtArrays =
Constants.sizeScales
|> List.map (\size -> ( size, Array.repeat size 0 ))
|> List.map (\( size, array ) -> ( toString size, \_ -> Array.append array array ))
|> scale "Hamt Array"
uint8Arrays : Benchmark
uint8Arrays =
Constants.sizeScales
|> List.map (\size -> ( size, JsUint8Array.repeat size 0 ))
|> List.map (\( size, array ) -> ( toString size, \_ -> JsTypedArray.append array array ))
|> scale "Uint8 Array"
Here are visuals with the plot and before the plot (that take roughly the same size as previously):
I don’t get much valuable info from these plots. Please let me know if you think my request is unjustified, and thanks for the great job for this library! It really is a pleasure to benchmark with it!
PS: Ultimately, a plot would be very nice, but not those that are presented. In my wildest dreams, for scales, elm-benchmark would automatically provide a plot, not of the samples, but of the timing performance for each scale (maybe a boxplot to encode standard deviation or fitting) with logarithmic scale, of the methods compared 
PPS EDIT: if this context of comparing more than two methods at different scales with a final plot is something you think is worth exploring, I’m willing to try to implement something.
