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.

本练习是课程的一部分

Data Manipulation in Julia

查看课程

练习说明

Group chocolates by company and iterate over the GroupedDataFrame.
Subset each group using ismissing() and the rating column, replacing the missing values by the value of replace_missing() function.

交互式实操练习

通过完成这段示例代码来试试这个练习。

# 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))

编辑并运行代码