Welcome to DL201 Discussions!

[jamescavanagh]

Here is a good place to ask questions.

[Jing-Pan-China]

Hi, I have a question about missing value filling. How can I fill the missing value of variable "Weight" using its average value based on "Team". For example, if "Team" finland miss some values of "Weight", I can use the average weight value of Finland to fill the missing value who is Finnish. The whole dataset is here: https://www.kaggle.com/heesoo37/120-years-of-olympic-history-athletes-and-results

[jamescavanagh]

Well, it really depends on the data. If there are other features which correlate with weight, such as height, you can make a model that will make a prediction for the weight based off of the height, ect.

In this case, it is a good first start.

Note, if most of the columns are also missing for that sample, then it would be a good idea to delete it

[Jing-Pan-China]

Hi, James. Thanks for your answer! I like your way to predict the weight based on the height.

But if I want to use the mean value about weight of a specific country to fill the missing value, I still do not know how to use pandas to finish it. I have already known that I can use mean value about weight of all countries to fill the missing value with the following code

.

[jamescavanagh]

Ok, in this case, it can be done using pandas.groupby() to take the mean height of each country, then mapping the mean onto the missing values

[jamescavanagh]

Step one would be using .group_by
https://chrisalbon.com/code/python/data_wrangling/pandas_apply_operations_to_groups/

to get a series with the names

step 2 would be use a map
https://chrisalbon.com/code/python/data_wrangling/pandas_map_values_to_values/