Closed boshek closed 6 years ago
Merging #46 into steffi_exp will decrease coverage by
0.66%
. The diff coverage is100%
.
@@ Coverage Diff @@
## steffi_exp #46 +/- ##
==============================================
- Coverage 93.86% 93.19% -0.67%
==============================================
Files 5 5
Lines 505 485 -20
==============================================
- Hits 474 452 -22
- Misses 31 33 +2
Impacted Files | Coverage Δ | |
---|---|---|
R/stations.R | 96.55% <100%> (-2.52%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update ccbabcc...e2cbded. Read the comment docs.
I don't think I'm happy with simply removing that option. I'd much rather make the sp package a suggests, then check for the package before using coordinates. Then if the user doesn't have it, they'll get a message saying to install sp. I'd also still like to look into a way of including this functionality, perhaps with another package. But haven't had the chance to look into it yet.
I would also +1 the suggests proposal to simplify the install but keep this feature.
…dependency and would close #45
Merging to
steffi_exp
as I think this is where you are doing much of the review work.This is an aggressive fix to the problem of importing the
sp
package. And it is a real shame because it is quite nice code. I would argue though that this reaches beyond the scope of the weathercan package and searching for stations by coordinates can be done with the sp (or preferably the sf) package as users zero in on exactly what stations they want. As I mentioned in the issue, I think importing sp is such a high cost for only this functionality.