diff --git a/_episodes_rmd/03-regression-regularisation.Rmd b/_episodes_rmd/03-regression-regularisation.Rmd
index 23ae7117..7e767e60 100644
--- a/_episodes_rmd/03-regression-regularisation.Rmd
+++ b/_episodes_rmd/03-regression-regularisation.Rmd
@@ -127,12 +127,12 @@ than observations.
 > visualise patterns when there are too many features!). 
 > 
 > ```{r corr-mat-meth, fig.cap="Cap", fig.alt="Alt"}
+library("ComplexHeatmap")
 > small <- methyl_mat[, 1:500]
 > cor_mat <- cor(small)
 > Heatmap(cor_mat,
 >     column_title = "Feature-feature correlation in methylation data",
 >     name = "Pearson correlation",
->     col = col,
 >     show_row_dend = FALSE, show_column_dend = FALSE,
 >     show_row_names = FALSE, show_column_names = FALSE
 > )
@@ -1049,8 +1049,7 @@ plot_elastic(0.75)
 >     update_role(age, new_role = "outcome") %>%
 >     ## center and scale all the predictors
 >     step_center(all_predictors()) %>%
->     step_scale(all_predictors()) %>%
->     prep(training = training(split_data), retain = TRUE)
+>     step_scale(all_predictors()) 
 > 
 > ## set the "engine" to be a linear model with tunable alpha and lambda
 > glmnet_model <- linear_reg(penalty = tune(), mixture = tune()) %>% 
@@ -1062,7 +1061,7 @@ plot_elastic(0.75)
 >     add_model(glmnet_model)
 > 
 > ## 5-fold cross-validation repeated 5 times
-> folds <- vfold_cv(training(split_data), v = 5, repetitions = 5)
+> folds <- vfold_cv(training(split_data), v = 5, repeats = 5)
 > 
 > ## define a grid of lambda and alpha parameters to search
 > glmn_set <- parameters(