Define series_types for pivoted measures meeting criteria?

[data2state]

I have a LookML dashboard file that I want to show bars for monthly values and lines for running totals. I know how to do this one-value-at-a-time in the graph through series->customization, and I also know you can set values by typing them out in “series_types:” parameter in LookML. The problem is that the data is pivoted for a large number of possible combinations, so the lines are reset to bars each time a user changes the filtering.

Specifically, the data is pivoted for each city and calendar year, and users can select multiple cities for any date range, so there are thousands of possible pivots. I was wondering if there was a way to make it so that “running_total” values were always plotted as a line, regardless of if they are for Phoenix 2012 or Dallas 2020 or Boston 2019 or whatever, without having to explicitly list all of those options in LookML or force the users to select the series customization each time, and while keeping the monthly values plotted as bars like default.

Any thoughts? I tried playing around with ${running_total}: Line format in the LookML but that didn’t have any effect, and simply using running_total: Line wouldn’t work because each series is named “City Year - running_total” as it is pivoted.

In researching, I came across this thread which mentions that Looker does not (or at least did not then) support a column order type customization: 

https://community.looker.com/dashboards-looks-7/custom-visualization-3845?postid=8176#post8176

Thank you for any help!