Visualise changes over time #175
Conversation
Also remove unused imports
Also converts data to array of objects if only a single object
Also fixes default classes for nodes and removed redundant code
|
Okay, had to hand roll my own BODS 0.3 test data (because examples are so broken): https://gist.github.com/radix0000/5e612d582b41a8aca5bfff4d5e396bf8 So doesn't show change over time with BODS 0.3 (which is fine if that is the intention), but does go into "change-over-time mode" with slider displayed if there is a replacesStatement property present (you just can't slide the slider, and all statements displayed), but latest time it shows conflicts with the fact all statements are shown, which is a bit confusing. So if it is intended that doesn't support change-over-time for 0.1-0.3 then recon that slider should always be hidden regardless of whether there is a replacesStatement property or not. |
|
@radix0000 Thank you. @kd-ods and I discussed this at length and came to the decision that 0.4 data renders and interactive slider, whereas pre-0.4 data with a change-over-time element renders an inactive slider to indicate that there is change-over-time data, but not render the changes themselves. The slider is hidden only if there is no change-over-time data detected.
I think this is an issue with the data. If you use this data, it works as expected. If you use your data in the current (old) visualiser, it renders much the same as here. |
Resolves comment #175 (comment)
codemacabre
force-pushed
the
ml/changes-over-time
branch
from
9d858cc to
3006086
Compare
There was a problem hiding this comment.
Looks good. One thing I happened to notice when testing is that data with the wrong version (i.e. bodsVersion == "0.4" when it is actually 0.3 format) results in a "Your data does not have any information that can be drawn." message, and while it should certainly give an error, ideally it could be more informative (especially as the only difference between data that works and doesn't, is one character). Possibly having a check for "recordDetails" property and whether its presence or not matches the version, < 0.4 vs >= 0.4, could work. I am doing something similar in the data review tool). But don't think it is a show stopper, so happy for this to be merged either way.
Note, this update introduces a bug with the arrowheads which will be addressed in a future issue (see #174).