Forum Replies Created
-
Posts
-
Hi Cristián,
I think it would be very helpful for a lot of us. I’ll try to come up with something soon.
Bill
Good question! As you’ve discovered, each organization of the data results in a different number of points in the two graphs. And only one organization at a time is possible. So, it is not possible to produce both a graph with 12 month points and a graph with 24 hour points at the same time.
A workaround which you may already have discovered is to click on the graph’s camera icon and open the image in the draw tool where it will be displayed statically, not changing when you restructure the data. (Of course you forgo dynamic selection for the static image.)
Bill
Hello Christián,
Thank you very much for this bug report. I have also seen this bug, but have had trouble reproducing it consistently. Your air quality document seems to provide the easy reproducibility we need to fix it.
Yours,
Bill
p.s. (April 2, 2018) The bug is now fixed and will appear in the next build—0439.
Hello Christián,
Generally, 5000 cases is the upper limit we use when we think about distributing CODAP documents, especially for classroom use.
Of course we are constantly working on improving performance, and would be very interested in egregious examples of slowness caused by something other than large numbers of cases.
Bill
I opened the document in a JSON editor and I can see that the assigned colors to all the Bioma categories is a gray. See screenshot enclosed. But you can use the brush palette in the graph to reassign them to nice colors (see second screenshot).
An important question is how this document corruption occurred.
Hi Christián,
Thanks for reporting this strange behavior. It’s a known bug in CODAP whose fix should appear in build 0438.
Bill
Hello Kathleen,
The trick here is to structure your data in a way that CODAP can use to display multiple box plots.
So, if you have an attribute for, let’s say Group, and another, let’s call it Measurement, your data might look like this:
Group Measurement
before 7
after 5
before 10
after 9
etc.The you put Measurement on one axis, and Group on the other.
Finally, in the Ruler menu for the graph click on the checkbox for Box Plot.
See the enclosed screenshot.
Hope this helps,
Bill
Attachments:
Hi Randy,
Yes, it is possible, and with a single formula (though perhaps not as elegant as one would wish). The screen shot shows how, using attributes x and fLW. It’s basically a nested ‘if/else’. Notice that there’s no last else clause so when x is 13, fLW gets no value.
if(x<=10, 6.6,
if(x=11, 10,
if(x=12,0)))Attachments:
Regarding state boundaries, let’s assume you have an attribute whose values are names of states. Let’s say the name of this attribute is State.
Now you need an attribute that whose values are the boundaries themselves. The name of this attribute can be any of the following: ‘boundary’, ‘boundaries’, ‘polygon’, ‘Boundary’, ‘Boundaries’, ‘Polygon’
Finally, you write a formula for this boundary attribute: lookupBoundary(US_state_boundaries,State)
After you accept this formula you should see for values small blue polygons that represent the boundaries for each of the states.
Hope this helps,
Bill
Hi Meggie,
Thanks for this question.
Regarding coordinates: CODAP looks for attributes with names that suggest latitude and longitude.
For latitude these are: ‘latitude’, ‘lat’, ‘Latitude’, ‘Lat’, ‘LAT’, ‘LATITUDE’, ‘LATITUD’, ‘latitud’
And for longitude these are: ‘longitude’, ‘long’, ‘lng’, ‘Longitude’, ‘Long’, ‘Lng’, ‘LONG’, ‘LON’,
'LONGITUDE', 'LONGITUD', 'longitud'
Once you have attribute names CODAP can recognize, making a map should show points for each case. (Close any maps you have open first.)
I’ll answer the question about boundaries in a following post.
Bill
Hi Mimi,
OK, now I understand. I think the right answer is that you should do it in a spreadsheet by appending a column to the Laundry dataset and filling it with “Laundry.” Then, when you import this into CODAP, you can drag that column to the left to create the parent “Location” collection.
But you inspired me to see if there was any way to do it from within CODAP. There is, but it’s very clumsy, and I don’t recommend it. But here are the steps, at least approximately:
- In the “top 2” dataset make a new attribute named “Time”
- Drag the case and location attributes to the left, leaving Time on the right.
- Click in the Index column of the child collection and insert as many cases as you have in the Laundry dataset.
- Give the Time attribute a formula like the one in the screenshot below, except referring to “Time” instead of “Temperature.”
This should have transferred all the temperature values into the top 2 dataset at the child level. Add the rest of the attributes, varying the formula to refer to each additional attribute in the Laundry dataset.
What a lot of work! Much easier in a spreadsheet.
There are a lot of data manipulation capabilities we would like to add to CODAP as we go along. “Ease of join” is certainly one of them!
Bill
Attachments:
Hi Mimi,
CODAP shows the “parent-child” relationship in the case table with curvy lines, as you can see in the screen shot I’ve uploaded. Very useful is to click on one of the parent rows to select that row and all the child case rows. This is also shown in the screen shot. As you scroll the child table, CODAP attempts to keep the parent table in synch so you can see the parents approximately next to their children.
Another way is to make a graph (as shown in the second screen shot) in which some unique identifier for the parent is on one axis and some attribute for the children is on the other axis. Here you can easily see that the selected child case has Kentucky as its parent.
I’m not sure I’ve answered your question. Don’t hesitate to follow up!
Bill
Hi Michelle,
Great suggestions. You inspired me to look carefully at the options offered by Leaflet, the library underlying CODAP’s map tool. I found ‘worldJump,’ which, when set to true, does part of what you describe. Namely, when you scroll the map horizontally far enough, say to the right, the map wraps and you start seeing points on the left that you had previously scrolled off to the right. This options lessens the likelihood that you’ll get “lost” and be unable to find your data. (Though note that the Rescale button at the top of the map’s tool palette scales the map so that all the data is visible.) Expect to see this enhancement in an upcoming release, perhaps 0412.
Unfortunately, the worldJump option still only allows you to see “one world” at a time, so you can still only see half the ring of fire at once. To help with this, consider using a formula for Longitude such as
if(Longitude_< cutPoint, 360+Longitude_, Longitude_)
where Longitude_ is the original longitude, and cutPoint is the value of a slider. Here is your example modified to include a cutPoint slider.
I’ll have to think a bit more about how hard it would be for us to force map points to automatically plot in whatever world is showing. It seems like it might not be too hard.
So, thanks for your excellent suggestions! You are helping to make CODAP a better educational platform.
Bill
Hi Kathy,
Bill, how do we limit the number of measurement tools? I noticed in the numbers plot on the right you allowed only five.
Do you mean how many “checkboxes” show in the Ruler menu. I’m not doing anything special and currently there is no way for you as the author to control what shows up there.
Bill
