Release 1.4-1 ?

[rsbivand]

Given #72, should we release sp again to calm CRAN a bit? Should we ask whether CRAN would like a pre-submission tarball to see if things are fixed? I'm running revdeps, but with NG rgdal and sf, so many positives don't apply to CRAN rgdal and sf.

[rsbivand]

Oops - in fasterize:

library(sf)
library(fasterize)
p1 <- rbind(c(-180,-20), c(-140,55), c(10, 0), c(-140,-60), c(-180,-20))
hole <- rbind(c(-150,-20), c(-100,-10), c(-110,20), c(-150,-20))
p1 <- list(p1, hole)
p2 <- list(rbind(c(-10,0), c(140,60), c(160,0), c(140,-55), c(-10,0)))
p3 <- list(rbind(c(-125,0), c(0,60), c(40,5), c(15,-45), c(-125,0)))
pols <- st_sf(value = rep(1,3),
              geometry = st_sfc(lapply(list(p1, p2, p3), st_polygon)))
r <- raster(pols, res = 1)

for my sf gives:

Browse[2]> attr(x[[attr(x, "sf_column")]], "crs")[["proj4string"]]
NULL
Browse[2]> CRS(NULL)
Error in if (!is.na(projargs) && !nzchar(projargs)) projargs <- NA_character_ : 
  missing value where TRUE/FALSE needed

I'll add a NULL trap to sp::CRS().

[rsbivand]

Done, will commit, push, and PR, as others may be affected. For fasterize, the example now fails on

> r <- fasterize(pols, r, field = "value", fun="sum")
Error in fasterize(pols, r, field = "value", fun = "sum") : 
  Index out of bounds: [index='proj4string'].
> debug(fasterize)
> r <- fasterize(pols, r, field = "value", fun="sum")
debugging in: fasterize(pols, r, field = "value", fun = "sum")
debug: {
    .Call("_fasterize_fasterize", PACKAGE = "fasterize", sf, 
        raster, field, fun, background, by)
}

because the Rcpp code accesses:

    Rcpp::CharacterVector sfproj4 =
      Rcpp::as<Rcpp::StringVector>(
        Rcpp::as<Rcpp::List>(polygons.attr("crs"))["proj4string"]
      );
    if(sfproj4[0] != NA_STRING) {
      Rcpp::S4 rcrs(raster1.slot("crs"));
      rcrs.slot("projargs") = sfproj4;
    }

without checking that the sf "crs" ojbect has a "proj4string" component. @noamross , @mdsumner ?

[edzer]

@noamross see also r-spatial/sf#1225

[edzer]

@rsbivand let me know when to release.

[rsbivand]

Maybe ask by reply to the CRAN email first, so that they know that it is coming? Roger Roger Bivand Falsensvei 32 5063 Bergen tir. 25. feb. 2020, 22.47 skrev Edzer Pebesma <notifications@github.com>:

…

@rsbivand <https://github.com/rsbivand> let me know when to release. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#73?email_source=notifications&email_token=ACNZ3BDYD5K74CTEAXINS63REWGX7A5CNFSM4K3C3JF2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEM5UKUI#issuecomment-591086929>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACNZ3BFFP7ZANI6NNNK2WVLREWGX7ANCNFSM4K3C3JFQ> .

[rsbivand]

and yes, I think the change addresses the issues seen on submission of 1.4-0 and found by Brian in raster. Roger Bivand Falsensvei 32 5063 Bergen tir. 25. feb. 2020, 22.50 skrev Roger Bivand <roger.bivand@gmail.com>:

…

Maybe ask by reply to the CRAN email first, so that they know that it is coming? Roger Roger Bivand Falsensvei 32 5063 Bergen tir. 25. feb. 2020, 22.47 skrev Edzer Pebesma ***@***.***>: > @rsbivand <https://github.com/rsbivand> let me know when to release. > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#73?email_source=notifications&email_token=ACNZ3BDYD5K74CTEAXINS63REWGX7A5CNFSM4K3C3JF2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEM5UKUI#issuecomment-591086929>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ACNZ3BFFP7ZANI6NNNK2WVLREWGX7ANCNFSM4K3C3JFQ> > . >