Filter on flexible arguments

[diazrenata]

Maybe I'm not seeing something.

I would like to be able to pass duckdbfs a set of arguments and have it perform filter operations using those, without having to spell out in advance what all those arguments will be.

Like this:

my_filter_fxn <- filter(my_data, filter_criteria){
    filter(my_data, filter_criteria)
}

behaving something like

my_filter_fxn(my_data, filter_criteria = c("coat == 'calico'"))

returning all rows of my_data with coat==calico. This is easy to do if you know to expect certain args, but I don't know how to do it if you don't know in advance whether the function is going to be called to filter based on coat, or on whisker length, or whatever.

[diazrenata]

(It is also not impossible to do using bracket subsetting, but I want to do this using duckdbfs which to my knowledge works with dplyr syntax).

[diazrenata]

https://dplyr.tidyverse.org/articles/programming.html#transforming-user-supplied-variables

looks like you could parse something.

[diazrenata]

https://stackoverflow.com/questions/24569154/use-variable-names-in-functions-of-dplyr :eyes:

[diazrenata]

filter_on_passed_vars <- function(df, conditions = list("NYEARS == 2",
                                                        "TREE_UNIQUE_ID == 27_3_65_20067_3_19")) {

  splitcond = lapply(conditions, strsplit, split = " ") 

  call1 <- call(splitcond[[2]][[1]][2], as.name(splitcond[[2]][[1]][1]),
                                         (splitcond[[2]][[1]][3]))
  calls <- lapply(splitcond, FUN = function(a_split_condition)  call(a_split_condition[[1]][2], as.name(a_split_condition[[1]][1]),
                                                                     (a_split_condition[[1]][3])))

  df |>
    mutate_all(as.character()) |>
    filter(!!!calls)
}

This works but there is a warning about the !!! eval syntax.

....but I think that's OK? https://rlang.r-lib.org/reference/splice-operator.html

TBD if this works with duckdbfs tho. It works with dplyr but IDK about within duckdb.

(surely other ppl do this with sql calls, so there's also that route...)

[diazrenata]

I have this working. https://github.com/diazrenata/in-the-trees/blob/main/R/query_tables.R

[diazrenata]

@Aariq Here is where I had notes on passing variable and function names to filter, etc. Have at it!

[Aariq]

You can do this with rlang like so:

f4 <- function(...) {
  exprs <- rlang::enquos(...)
  iris |> 
    dplyr::filter(!!!exprs)
}

f4(Species == 'setosa', Petal.Width < 0.3, Sepal.Length == 5.1)
#>   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
#> 1          5.1         3.5          1.4         0.2  setosa
#> 2          5.1         3.4          1.5         0.2  setosa
#> 3          5.1         3.8          1.6         0.2  setosa

^{Created on 2024-01-31 with reprex v2.0.2}

Not sure how to get this working with a list as an input, only with ...

[diazrenata]

Cool, I'll see if this can work!

On Wed, Jan 31, 2024 at 4:43 PM Eric R. Scott @.***> wrote:

External Email

You can do this with rlang like so:

f4 <- function(...) { exprs <- rlang::enquos(...) iris |> dplyr::filter(!!!exprs) }

f4(Species == 'setosa', Petal.Width < 0.3, Sepal.Length == 5.1)#> Sepal.Length Sepal.Width Petal.Length Petal.Width Species#> 1 5.1 3.5 1.4 0.2 setosa#> 2 5.1 3.4 1.5 0.2 setosa#> 3 5.1 3.8 1.6 0.2 setosa

Created on 2024-01-31 with reprex v2.0.2 https://reprex.tidyverse.org

Not sure how to get this working with a list as an input, only with ...

— Reply to this email directly, view it on GitHub https://github.com/mekevans/forestTIME/issues/13#issuecomment-1920187438, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEH6DN6UEVOPUZJDYR4LJYLYRLJLFAVCNFSM6AAAAABBTYE2I6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRQGE4DONBTHA . You are receiving this because you modified the open/close state.Message ID: @.***>