Review notebook 6 "Calculate statistics" #8
Conversation
|
Check out this pull request on You'll be able to see Jupyter notebook diff and discuss changes. Powered by ReviewNB. |
| @@ -0,0 +1,873 @@ | |||
| { | |||
There was a problem hiding this comment.
This is an excellent attempt at explaining groupby with a diagram! However, I think it would be easier to understand if there were only 2 groups in the diagram, rather than 3. Part of the reason is that the 4 shades of green are difficult to distinguish between.
Reply via ReviewNB
| @@ -0,0 +1,873 @@ | |||
| { | |||
There was a problem hiding this comment.
I didn't quite follow as to why you created titanic_subset and did a groupby on it, and then separately did a groupby on titanic, even though they have the same result (except for the data type of the result). I think there is a teaching point you were trying to emphasize by making that distinction, but I couldn't quite figure it out!
Reply via ReviewNB
There was a problem hiding this comment.
Thanks for noticing, I'll try to improve the _storyline_ . The reason to have the subset is to overcome the necessity of selecting a column (e.g. Age) after the groupby operator. As such, I can first explain
titanic.groupby("Sex").mean() before doing titanic.groupby("Sex")["Age"].mean().
Maybe rather work without subset, just do titanic.groupby("Sex").mean(), get a grouped-average result on all the columns (PassengerId,Survived,Pclass,Age,SibSp,Parch and Fare) and next, explain how to just go for a single column of interest?
There was a problem hiding this comment.
After trying out, I rather stick to:
titanic[["Sex", "Age"]].groupby("Sex").mean() as the solution of the specific question. It is putting less attention to the subset as it was before, but still keep the focus on those two columns involved.
| @@ -0,0 +1,873 @@ | |||
| { | |||
There was a problem hiding this comment.
No description provided.