for folks using R studio, if you want info about a function, put the cursor on the function and hit the F1 key. It does the same thing as ?my_func in the console but is faster if you have already included the function in your code.
Very interesting. I do still think that in this dataset little is gained by imputation. With so few missings it has practically no effect. However, the exercise itself is of great value because it demonstrates how (and perhaps when) to impute different types of missing values. I would like to see more on imputation and perhaps a bit of dive into the different types of missings (MAR, MCAR and MNAR) which boggle the mind. Thanks for another great video.
Me too! Methods like mice, missRanger, step_knn, missMDA, using PCA and all that stuff. :) There are so many alternatives, that I think it would be nice to compare each method in an easy way. Probably each dataset benefit differently with regards to the imputation method
@@EquitableEquations true, forgot that. after some considerations and if else should do the trick when the value is either 0 or 1 each day, otherwise i guess a fill in both directions
You can find materials supporting this vid (and others) at github.com/equitable-equations/youtube.
for folks using R studio, if you want info about a function, put the cursor on the function and hit the F1 key. It does the same thing as ?my_func in the console but is faster if you have already included the function in your code.
Very interesting. I do still think that in this dataset little is gained by imputation. With so few missings it has practically no effect. However, the exercise itself is of great value because it demonstrates how (and perhaps when) to impute different types of missing values. I would like to see more on imputation and perhaps a bit of dive into the different types of missings (MAR, MCAR and MNAR) which boggle the mind. Thanks for another great video.
Me too! Methods like mice, missRanger, step_knn, missMDA, using PCA and all that stuff. :) There are so many alternatives, that I think it would be nice to compare each method in an easy way. Probably each dataset benefit differently with regards to the imputation method
❤
for working_day I would sort the data (already done) then group_by day and then tidyr::fill and then ungroup
Yep! That works great unless midnight is missing, in which case the wrong day gets filled 🤓
@@EquitableEquations true, forgot that. after some considerations and if else should do the trick when the value is either 0 or 1 each day, otherwise i guess a fill in both directions