Thanx guys,
practical and effective solutions.
Gianandrea
glaporta wrote:
>
> Hi,
> sqldf is a fantastic package, but when the SELECT procedure runs unused
> levels remain in the output. I tried with the drop function, but without
> success. Do you have any suggestions?
> Thanx, Gianandrea
>
>
Here are two approaches:
1. rename the column to circumvent the column classing heuristic. In
the following
Species will be a 3-level factor and Species2 will be a 2-level factor
sqldf("select *, Species as Species2 from iris where Species <> 'setosa'")
2. use method = "raw". This code below r
If you want to suppress the unused level of Species, you can use factor() :
table(base$Species)
table(factor(base$Species))
2008/8/28 glaporta <[EMAIL PROTECTED]>
>
> Hi,
> sqldf is a fantastic package, but when the SELECT procedure runs unused
> levels remain in the output. I tried with the dro
3 matches
Mail list logo