Problem with some numbers in Norwich about pct-shiny HOT 6 CLOSED

Was assigned to me as thought it was data issue. Now I don't think it is... I've reproduced it for Manchester also: when you select SIC values the attributes of the line popup seem to change compared with SLC and OLC. @nikolai-b and @usr110 can you reproduce? It's a strange issue, may be caused by re-ordering and I think it applies to all cities.

from pct-shiny.

I can't reproduce, is this fixed for you @Robinlovelace ?

from pct-shiny.

@nikolai-b , it's definitely a bug unfortunately. In the above example you are changing the zone attribute. But this is a line-based bug! So 'Line attribute to display' is what should be changing. I've just generated 2 reproducible examples. I suspect that a fix is related to the popup or the ordering. A tricky one - dangerous as it went so long undetected. For a mass of line in east Manchester:

And for the furthest point south on the map, isolating one line, showing the entire screen so you can see my settings:

from pct-shiny.

Hey Robin,

After a bit of digging here is what I found:
Picking a random line in MCR E02001065 E02001067

> grep('E02001065 E02001067', l@data$id)
820

but..

> grep('E02001067 E02001065', l@data$id) 
920

and shock horror:

> l@data$All[820]
43
> l@data$All[920]
30

So, I guess this is kind of data issue but maybe not.
If we added flow lines then it would be fine to keep it as it is
(becuase then it would be clear when the number changed that
it was not the same data but same line, different direction). Thoughts?

from pct-shiny.

Hi Nikolai, this is one to talk about.

I agree - direction may be an issue: 'E02001065 E02001067' is the flow from MSOA E02001065 to E02001067 and the other is in the other direction. Perhaps when generating the lines I need to aggregate these flows - probably a good shout for size, simplicity and robustness reasons. Strange that it only happens for SIC though? Or perhaps its just the order it's plotting the lines that's the issue. Well spotted, cheers for reproducing this and no thanks for giving me more data manipulation work :)

from pct-shiny.

It happens only in SIC because that is where the reverse flow is shown to be 'top', why that is the case is more a question for you 😄. Aggregating is defiantly is the easiest solution for me to implement! I will close this as it has become a data issue.

from pct-shiny.