[Feature] Sum up allocation sub classes

[putzwasser]

Currently fava_investor shows the percentage for each asset class in relation to the total:

It would be nice if fava_investor showed the sum of a class in regards to the total and to its parent(s):

(Something like this)

This might get confusing if there are more than three sublevels. Therefore, It might be best to include an addition column:

• Ratio in parent which holds the ratio in regards to the total of the parent.

[redstreet]

It already does: hover your mouse over the percentage. Give it a shot and let me know.

[putzwasser]

Ah, I see 👍

May I suggest some improvements:

• Displaying the amounts/percentages on hover is unintuitive and people probably will never/not know this piece of informatin exists (I'm the best example 😆 )
• Having to hover over the rows to see the data makes it very difficult to compare allocations, e.g., to see if you're off your target allocation.
• Therefore, I'd favor a solution where the data is either always present or visible for all rows after a single click (key is best UX, IMHO).*

* I noticed that the data shows if you collapse the "subrows", but then again, part of the information is missing.

Say, your target domestic stock allocation is 45%. First, you need to collapse the domestic row. Then you'd see your current allocation is 50.12%. Then, you'd be wondering: “Which position would I want to decrase?” One more click to expand domestic again, to see the allocation within it. “Is largecap as big within domestic as I think?” => Hover above largecap.

I mean, this works. But it is cumbersome in my opinion.

Please don't get me wrong. I don't want to be ungrateful for the work you've done. I really enjoy it. Just wanting to contribute to better data accessibility/visibility.

[redstreet]

I agree that these types of things in general would be a nice to have. Representing hierarchical percentages well in a GUI to be useful to 90%+ of the users can be a bit of a challenge. The way people use AA is very individual, and so, this needs to be thought out well through a broader set of use cases.

BTW: I just opened #78 for related higher level functionality.

There are many good ideas here, and I'm all for experimenting with them, but the bottleneck is contribution time. Volunteer contributions would help, since I'm unlikely to get to this soon. Feel free to submit a patch.