Would it be possible to lock/save chart filter settings so that they don’t reset every time you switch between present and model modes? I have some charts in which I want to display only certain category elements, and every time I switch views I lose the display filter.
Great idea! We could include functionality to lock which category items to show in Present mode, then next time you go back to Present mode, that filter will persist
I think that’ll be a great improvement to the UI. and it’s the same thing for tables where you set a filter rather than wanting to see the aggregate view. It’ll be really helpful
Yes, it reminds me of the value of creating multiple views of the same list in SharePoint.
Indeed - I would find that particularly useful for variables with multiple layers of categories, where I want to slice and dice the data in different ways in the reports, but need to maintain the granularity in the variable itself
Speaking of multiple layers, I really like being able to apply multi-dimensional accounting where any given account or line item can be associated with multiple tags, and each tag is in its own layer of one of many standard taxonomies, but that may be getting more sophisticated in the early years of Causal.
Could you give us some examples of the different tags an account/line item might be associated with? And then how you might like to see those different tags/groups of tags presented?
You’re correct that it probably wouldn’t be a short-term feature, but in the longer term we are thinking about how to make Causal more flexible
I gave the example of dimensions in accounting, like what various accounting programs can do for reporting of different sets of books like the set of books for taxes vs the set of books for GAAP vs the set of books for non-GAAP financial statements. Dimensions, Sage Intacct Reporting Dimensions | Sage Intacct, Inc.
But in modeling a commercial real estate or M&A deal, it would be useful to create multiple hierarchies of dimensions. So instead of separate dimensions for departments, business units, groups, divisions, subsidiaries, etc. there would just be one hierarchy of the corporate structure. Then if I chose to tag the business unit, the levels above it would also become associated to different degrees whereas the departments wouldn’t report on variables or constants tied to the tag of a particular business unit or all business units in general.
Another hierarchy of dimensions could be location - not only am I in Minneapolis, but I’m also in Hennepin County, within the Metropolitan Council’s jurisdiction, and the regional Minneapolis/St Paul MN-WI MSA, the state of Minnesota, the USA, North America, and the Americas.But maybe I don’t want it tied to the zip code, ward, precinct, or whatever within Minneapolis, so I just tag Minneapolis.
Another dimension could be a customized set of portfolios > programs > projects; a hierarchy of a particular supply chain; the layers or taxonomy of relevant industry codes for market analysis, etc.
I would think the main tags of each hierarchy would be in little boxes underneath the name of the variable, constant, filter, etc. Then when I hover my mouse over it, then the hierarchy of that tag overlays on top where I can see the parent, grandparent, great grandparent, etc. going straight up, and see a list of the first layer of children underneath straight down in one column. Then I could click on any one of them to change which one of the tags is highlighted or active and then the lists above and below the tag update. This could be done with each tag all within the regular spreadsheet interface/view and not require any popup windows. Once my mouse moves away then the hierarchy disappears and all the tags probably go gray when my mouse isn’t anywhere near that variable/constant/filter.
I like the way you think Michael. I’m not coming from an accounting perspective, but I 100% agree that this would be a great functionality to add. This could also be applied to blocks of logic - at the moment we have sections and groups, so 2 levels of hierarchy, but what would be very cool would be to be able to have more than 2 layers of logic - taking your portfolios > programs > projects example, if you wanted to model and aggregate data across every layer, of if it was business unit > region > country > continent (for big companies). You can do it with models serving as aggregators, but the cool thing with having it all in one model, is you can templatize model logic and use it over and over.
You could also have metadata on variables - like a flagging system - so that you could say that this line item/variable applies to Minnesota, or USA or whatever. It’s a bit of a tangent, but if you could also use the names of the groups/sections, or metadata in the formulae, you could do some very cool stuff with aggregation/filtering in formulae - things like sum(variables in the section below me with the Minnesota flag), or sum(all variables below with the name Revenue).
It might take away for the current user-friendly way of writing formulae in Causal though, so I’m not sure if people would go for it.