Replacing rating with group median
In the last exercise, you replaced the missing values in the rating column with the column median. But could you do better? Yes! You can replace the missing values with the median rating of chocolates from the same company. Let's do it!
There is a predefined replace_missing() function that takes two arguments - a DataFrame group and a column col. It tries to compute a median of the column col and returns it if it is successful. If calculating the median fails, for example, because there are no values, then it returns a predefined value.
The chocolates dataset and the DataFrames and Statistics packages have been loaded for you.
Questo esercizio fa parte del corso
Data Manipulation in Julia
Istruzioni dell'esercizio
- Group
chocolatesbycompanyand iterate over the GroupedDataFrame. - Subset each group using
ismissing()and theratingcolumn, replacing the missing values by the value ofreplace_missing()function.
Esercizio pratico interattivo
Prova a risolvere questo esercizio completando il codice di esempio.
# Group by company and iterate
for group in ____(____)
# Subset each group using ismissing() and the rating column, assign a new value
group[____, ____] .= replace_missing(group, :rating)
end
println(describe(chocolates, :nmissing))