Fix Unit Tests for tuning SSM, tune_on_folds & nonignorable

[PhilippBach]

Currently, I can't see why the unit tests for tuning SSM models if setting tune_on_folds=TRUE and score="nonignorable" are failing. Errors are produced when calling initiate_learner` and the error comes from

doubleml-for-r/R/helper.R

Lines 240 to 262 in c7edeb2

	initiate_learner = function(learner, task_type, params, return_train_preds = FALSE) {
	valid_task_type = c("regr", "classif")
	assertChoice(task_type, valid_task_type)
	ml_learner = learner$clone()
	if (!is.null(params)) {
	ml_learner$param_set$values = insert_named(
	ml_learner$param_set$values,
	params)
	} # else if (is.null(params) | length(params) == 0) {
	# message("No parameters provided for learners. Default values are used.")
	# }
	if (task_type == "classif") {
	ml_learner$predict_type = "prob"
	}
	if (return_train_preds) {
	ml_learner$predict_sets = c("test", "train")
	}
	return(ml_learner)
	}

The error message is

Error in insert_named(ml_learner$param_set$values, params) : 
  Assertion on 'names(y)' failed: Must have names.

see #217

[PhilippBach]

For now I skipped the tests, see b0dd406

[petronelaj]

@PhilippBach The problem was that get_params() returnes params for all folds in a list but nonignorable needs just the one for the current fold.
Changing this

ml_learner_pi_hat = initiate_learner(
            learner = self$learner$ml_pi,
            task_type = private$task_type$ml_pi,
            params = self$get_params("ml_pi"),
            return_train_preds = FALSE)

to

if(length(self$get_params("ml_pi"))  == self$n_folds) {
            params_pi = self$get_params("ml_pi")[[i_fold]]
          } else {
            params_pi = self$get_params("ml_pi")
          }

ml_learner_pi_hat = initiate_learner(
            learner = self$learner$ml_pi,
            task_type = private$task_type$ml_pi,
            params = params_pi,
            return_train_preds = FALSE)

for all learners seems to fix it.