revdeps fail after froll Inf handling change

[tdhock]

two revdeps are affected by #7264

bidask https://rcdata.nau.edu/genomic-ml/data.table-revdeps/analyze/2025-08-30/bidask.txt

-* checking examples ... OK
+* checking examples ... ERROR
+Running examples in 'bidask-Ex.R' failed
+The error most likely occurred in:
+
+> base::assign(".ptime", proc.time(), pos = "CheckExEnv")
+> ### Name: edge_expanding
+> ### Title: Expanding Estimates of Bid-Ask Spreads from Open, High, Low, and
+> ###   Close Prices
+> ### Aliases: edge_expanding
+> 
+> ### ** Examples
+> 
+> # reduce number of threads to pass CRAN checks (you can ignore this)
+> data.table::setDTthreads(1)
+> 
+> # simulate open, high, low, and close prices with spread 1%
+> x <- sim(n = 1000, spread = 0.01)
+> 
+> # estimate the spread using an expanding window
+> s <- edge_expanding(x$Open, x$High, x$Low, x$Close)
+Error in froll(fun = fun, x = x, n = n, fill = fill, algo = algo, align = align,  : 
+  n must be positive integer values (> 0)
+Calls: edge_expanding ... EDGE -> rmean -> rfun -> froll -> frollfun -> froll
+Execution halted
 * checking for unstated dependencies in 'tests' ... OK
 * checking tests ... OK
   Running 'testthat.R'
 * checking for unstated dependencies in vignettes ... OK
 * checking package vignettes ... OK
-* checking re-building of vignette outputs ... OK
+* checking re-building of vignette outputs ... ERROR
+Error(s) in re-building vignettes:
+  ...
+--- re-building 'bidask.Rmd' using rmarkdown
+
+Quitting from lines 72-79 [unnamed-chunk-6] (bidask.Rmd)
+Error: processing vignette 'bidask.Rmd' failed with diagnostics:
+i In argument: `EDGE (expanding) = edge_expanding(open, high, low,
+  close)`.
+i In group 1: `symbol = "BTC"`.
+Caused by error in `froll()`:
+! n must be positive integer values (> 0)
+--- failed re-building 'bidask.Rmd'
+
+SUMMARY: processing the following file failed:
+  'bidask.Rmd'
+
+Error: Vignette re-building failed.
+Execution halted
+
 * checking PDF version of manual ... OK
 * DONE
-Status: OK
+Status: 2 ERRORs

timeplyr https://rcdata.nau.edu/genomic-ml/data.table-revdeps/analyze/2025-08-30/timeplyr.txt (output too large)

[jangorecki]

timeplyr failure is related to breaking change

  -- Failure ('test-roll_mean.R:14:3'): Expect NA --------------------------------
  lapply(x, function(y) roll_sum(y, na.rm = FALSE)) (`actual`) not identical to lapply(1:length(x), function(x) NA_real_) (`expected`).
  
    `actual[[5]]`: Inf
  `expected[[5]]`:  NA
  
    `actual[[6]]`: -Inf
  `expected[[6]]`:   NA

I will report that to maintainer asking for adapting to change

[jangorecki]

bidask issue seems to be related to randomly generated input.
I tested CRAN's bidask version and our devel data.table and it worked fine.
We should look into revdeps check results after next run, if we are still getting an error.
Reported the problem to bidask maintainer.

[ben-schwen]

@jangorecki We also have here a strange mismatch between adaptive=TRUE and single n values.

frollmean(1:2, n=c(0,0), adaptive=TRUE)
#> NaN NaN
frollmean(1:2, n=c(-1,0), adaptive=TRUE)
#> NA NaN
frollmean(1:2, n=0)
#> Error in froll(fun = "mean", x = x, n = n, fill = fill, algo = algo, align = align,  :
#>   n must be positive integer values (> 0)

[jangorecki]

@jangorecki We also have here a strange mismatch between adaptive=TRUE and single n values.

frollmean(1:2, n=c(0,0), adaptive=TRUE)
#> NaN NaN
frollmean(1:2, n=c(-1,0), adaptive=TRUE)
#> NA NaN
frollmean(1:2, n=0)
#> Error in froll(fun = "mean", x = x, n = n, fill = fill, algo = algo, align = align, :
#> n must be positive integer values (> 0)

Good spot, now I checked more in detail and bidask randomly calls adaptive functions depending on the input length
https://github.com/eguidotti/bidask/blob/d642e02055bc20843deb064d1ab6e5fd77954110/r/R/utils.R#L29

Adaptive n validation is definitely something to improve here it has been already improved and that's why it is throwing error now.

[jangorecki]

@jangorecki We also have here a strange mismatch between adaptive=TRUE and single n values.

frollmean(1:2, n=c(0,0), adaptive=TRUE)
#> NaN NaN
frollmean(1:2, n=c(-1,0), adaptive=TRUE)
#> NA NaN
frollmean(1:2, n=0)
#> Error in froll(fun = "mean", x = x, n = n, fill = fill, algo = algo, align = align, :
#> n must be positive integer values (> 0)

this mismatch occurred in CRAN data.table, in devel data.table is has been already fixed.

[jangorecki]

As of current master, bidask package should no longer be broken by data.table.
timeplyr maintainer said he will look into it this week. Considering this breaking change was expected we could possibly close this issue already, but let's keep it open to confirm downstream packages are working fine with devel data.table.

[jangorecki]

Yes, bidask breaking is resolved as per https://rcdata.nau.edu/genomic-ml/data.table-revdeps/analyze/2025-09-08/

[tdhock]

great thanks Jan