R mlr3 create TaskregrST duplicate rows?

Question

I have a dataframe called tab_mlr with coordinates about 19 features in 788 rows.

str(tab_mlr)

This object have 788 observations of 21 variables (with 2 variables as Latitude and Longitude). I create an sf object like this :

data_mlr <- sf::st_as_sf(tab_mlr, coords = c("Longitude", "Latitude"), crs = 4326)

data_mlr have 788 features, that's ok. But when i create a task with this data_mlr like this :

task <- TaskRegrST$new(
  "mlr",
  backend = data_mlr,
  target = "Hauteur"
)

task object have 620 944 rows !!! Why not 788 rows ?

If you can reproduce this with an example task please open an issue in the GH repo. — pat-s, Jun 02 '21 at 15:05

score 0 · Answer 1 · answered Jun 02 '21 at 15:03

0

The reason might be that you are making 788 rows for every row, and you have 788^2 rows as a result.

answered Jun 02 '21 at 15:03

Shivam

Yes, 788^2 = 620 944, but why have we 788 rows for every rows and not 1row for every rows ? – Pascal Obstétar Jun 03 '21 at 07:36

1 Answers1