.SDcols gives a strange message if length() 0 #3185

MichaelChirico · 2018-12-05T08:03:59Z

Also identified in a comment by @HughParsonage here:

For now I'll solve this by just short-circuiting to null.data.table() if !length(.SDcols)... the broader issue raised in #1789 of handling length-0 .SDcols in the case of := will require a bit more work.

MichaelChirico pushed a commit that referenced this issue Dec 5, 2018

Closes #1878 and #3185 -- .SDcols accepts patterns

b24f490

MichaelChirico pushed a commit that referenced this issue Dec 5, 2018

NEWS item for #3185

Loading status checks…

87a7bb6

MichaelChirico mentioned this issue Dec 5, 2018

RFC: .SDcols=patterns() #3186

Merged

mattdowle added this to the 1.12.0 milestone Dec 14, 2018

mattdowle closed this in #3186 Dec 14, 2018

mattdowle mentioned this issue Dec 20, 2018

Empty .SDcols back to 1.11.8 behaviour, plus fix for integer() #3238

Merged

Rdatatable / data.table

.SDcols gives a strange message if length() 0 #3185

.SDcols gives a strange message if length() 0 #3185

MichaelChirico commented Dec 5, 2018 •

edited

Rdatatable / data.table

Join GitHub today

.SDcols gives a strange message if length() 0 #3185

.SDcols gives a strange message if length() 0 #3185

Comments

MichaelChirico commented Dec 5, 2018 • edited

Essential cookies

Always active

Analytics cookies

MichaelChirico commented Dec 5, 2018 •

edited