Forum Replies Created
-
Posts
-
Hi Andee,
Yes, that’s correct, the new order will apply everywhere from then on (until you change it again). In fact, if you have two graphs that use the same categorical attribute and you reorder in one you will see the reordering take place dynamically in the other as well.
Bill
Hi Jo,
I second Dan’s request for a share link. I tried to duplicate your problem and came up with this document. It has a slider that dynamically affects the value of the attribute group.
Bill
Hi Jo,
I believe you have uncovered a new bug! I thought that all you had to do is to make a new slider and rename it to have the name of the lost slider. But this doesn’t work. It should.
So, what you have to do is create a new slider and name it differently than the lost slider, say “age_filter” (with the underscore character). Now modify the formula (or formulas) that refer to “agefilter” to refer to “age_filter” instead.
Everything should be fine again.
Bill
Hi Jo,
Good question!
The very first time you use a categorical attribute in a graph, it will be ordered alphabetically. After that, there is no automatic way to resort the categories, so you have to do it by hand. The ordering of the categories becomes part of the attribute itself.
Bill
Ah, I’m guessing from your screen shot that you’d like to display the units of the attribute?
No problem! In Edit Attribute Properties, you can specify the units, and these will be displayed very nicely in parentheses in the column header and on graph axes. Furthermore, you won’t have to worry about them in any formulas you write that refer to an attribute.
Hi Sue,
Text between angle brackets is stripped out of attribute names when they are displayed in the case table. But if you bring up Edit Attribute Properties you’ll see that the angle brackets and text between them is still there.
Any text before the ‘<‘ character and after the ‘>’ is displayed as normal.
Angle brackets in HTML are used as part of the markup and I believe they are triggering something in the case table display that we could probably get around. But I would recommend avoiding their use!
Can you be more specific about the problem with $ signs? I couldn’t get anything strange to occur with them.
Bill
I agree we need much better documentation on how to get boundaries in maps. We’re very close to having maps able to display multiple layers of points and boundaries together in the same map. That will definitely require new (and better) documentation.
Thanks for your feedback.
Bill
Hi Andrew,
In addition to the boundary attribute that contains the formula, you’ll need another attribute for the names of the states. Let’s say you called that attribute state_name. Then your formula for the boundary would be:
LookupBoundary(US_state_boundaries, state_name)
Then enter the names (or abbreviations) of states in the cells in the state_name attribute column of the table.
If you already have a map open, you may need to close it and open another one after you’ve got your attributes defined. (I’m working on a fix for this right now!)
Bill
Hello Andrew,
Your post relates to a powerful and interesting capability of CODAP, one that has been used by a number of projects.
The Map Data example pulls in state boundaries with a formula:
lookupBoundary(US_state_boundaries, state)
There are four other sets of boundaries available by these means: US_congressional_districts, US_county_boundaries, US_puma_boundaries, and country_boundaries. You can see documentation for how to do this in the formula editor under lookupBoundary as shown in the screenshot I’ve enclosed.
If an attribute is named one of ‘boundary’, ‘boundaries’, ‘polygon’, ‘polygons’ (case independent), CODAP will attempt to use it to plot boundaries in a map. If the type of an attribute is boundary as specified using attribute properties, it won’t matter what the name is.
The boundaries themselves are expressed in geojson which you can read about on the web. You can actually paste geojson into a value and have it display a boundary on a map.
There are online utilities for converting SHP files to geojson. Then you can export these to ‘csv’ and import them into CODAP.
I hope this helps you get started!
Bill
Attachments:
Hi Heinrich,
You have definitely revealed a bug, but there is a nice workaround. Put the days on the y-axis instead of the x-axis. You can do this on a second graph if you like, and as you drag the days in this second graph, they will also reorder in the original graph.
Meanwhile, I’ll log the bug.
Thanks,
Bill
Hi Sue,
Glad you got that to work! The sample attribute is typed as “categorical” just for this purpose.
Don’t hesitate to ask for help with the Sampler.
Bill
Hi Andee,
Excellent point, and I’ll log it as a story.
You probably know this, but you can change the color corresponding to the category that has been assigned gray using the ‘brush’ menu on the graph inspector. (But you can’t change color assigned to missing values.)
Bill
Hi Andee,
It would be very nice to be able to combine categories using a graph. (TinkerPlots gives you a way of doing this.) I’ll add it as a feature request.
You don’t have to write a formula, though. As shown in the screen capture, you can use the case table to group by the attribute that has the categories you want to combine. Create a new attribute in which you copy the existing category names, except where you want to combine.
Hope that helps!
Bill
Attachments:
