[kristenpblair@gmail.com]

Hi,

I am wondering if it is possible to control the scale of the colors displayed when an attribute is dragged onto a map. It seems like there can be 5 bins only, and what values go into each of those color bins depends on the spread of the data. I am hoping to get more granularity, and also to be able to keep the color scale the same between data displayed for different years, which have different variability. I can get the min and max to be the same by adding a row with the same min value and one with the same max value to each attribute column, but the values that are grouped into each color scale bin seem to change depending on the data. Is it possible to have more control over this scale? When I click on the ruler icon on the map, it doesn’t provide any options.

Thanks!

[Bill Finzer]

Hi Kristen,

The short answer is that CODAP does not provide any way to adjust the numeric legend color scale. As far as I know you are the first person to express a desire for this. Four years ago when we implemented the feature we thought users might want to adjust things but decided to wait and see. Now I’ll certainly add your feature request to our list.

Perhaps you could say a bit about the context in which you are using CODAP and how it would help you to make these adjustments?

Thank you!

[Jacob Sagrans]

I too would like to be able to do this. Look at the following CODAP document I prepared, for example. The banding seems really weird. Like there is one for 21 to 26 cases and another for 140 to 1566 cases. It may have been better to do this as a rate per 100,000 population or something like rather than total numbers but even then I might want to change the range each color represents.

https://codap.concord.org/app/#shared=158681

[Bill Finzer]

Hello Jacob,

Thanks for letting us know that you also would appreciate being able to change the ranges of the bands.

The bands are currently chosen such that each has approximately 20% of the values. This frequently leads to oddness, especially with highly skew distributions.

Our plan is to make it possible to drag the boundaries between bands on the legend and to make it possible add or delete bands.

Bill

[Jacob Sagrans]

I’m revisiting this topic now because for a project I am working on I took a CODAP file a colleague created with Lyme disease data over the past 15 some years, added a map to it, and then created a slider to show change in Lyme rates from year to year (Bill, I am really indebted here to what you did for the smallpox over the years CODAP file). See the file I made here. I noticed that the minimum and maximum Lyme rate values of course change from year to year, which is a little confusing. I wonder if there might be some advantage to instead displaying something like simply 1, 2, 3, 4, 5 to represent the different quintiles on the color legend? Or I wonder if the quintiles could be set based on the full dataset and not the subset for each individual years (although there may be some cons to this if there is a lot of variation from year to year)? Just thinking/wondering aloud here, but I do think this issue becomes more complicated when instead of a static map there is a slider showing change over time.

[Bill Finzer]

Hello Jacob,

I agree that the choropleth scale doesn’t play well with changing the subset of data one is viewing. Unfortunately, giving the user the level of control required to base the colors on the full dataset is difficult. But we’ll think about it!

Bill

[Author]

Posts