npct / pct-shiny

The Shiny map for Local Authorites
GNU Affero General Public License v3.0
24 stars 14 forks source link

Display increase in percentage for scenarios for route-network #358

Closed usr110 closed 8 years ago

usr110 commented 8 years ago

Also add in a caveat that further information can be seen in the model output tab

Robinlovelace commented 8 years ago

I suggest this is more a 'for the future' and discussion issue so closing for now. Would you mind suggesting it here https://github.com/npct/pct-shiny/issues/273 to keep the bug tracker for things that are definitely agreed and actionable.

Robinlovelace commented 8 years ago

Have agreed on this? One for discussion I think.

JDWoodcock commented 8 years ago

Please dont close issues without checking. Anna & I would like to see this down before launch. The reality is as we dont include all commuters in the route network the ratio is more accurate than the absolute nos.

Robinlovelace commented 8 years ago

OK fair point. Please discuss issues and get a check on effort, who can do it etc before opening next time... One to discuss tomorrow but agree an indication of % increase could be useful and probably not too hard to calculate in rnet.Rds and to display in the associated pop-ups.

Robinlovelace commented 8 years ago

Fast work @usr110 !

usr110 commented 8 years ago

Updated text for the route-network:

Between-zone cyclists* (baseline):
Between-zone cyclists* (scenario):
Ratio (scenario/baseline):

  • Selected cyclists: see Model Output tab
Robinlovelace commented 8 years ago

Hi Ali and all I think this is great work but think we need to modify the popup text to make it easier to understand (hence re-opening for now, even though the central issue has been resolved). Here's a concrete example from the latest London data: screenshot from 2016-08-11 16-31-13

The problem as I see it is that the meaning of "Ratio (scenario / baseline): 2.45" is not instantly clear, especially if you don't have formal technical training or an A Level in Mathematics.

Suggestion:

"Percent increase: 145%"

The code could look something like

round(100 * rnet$senario / rnet$bicycle - 1)

In this case that would look a bit like this for output:

> (100 * (12000 / 5000 - 1))
[1] 140

Sound good?

Also suggested change to the text:

Of trips in the model (see model output tab).

Robinlovelace commented 8 years ago

I think 'cyclists' is not the right word. Could also use 'commuter', but not generalisable, so suggest trips may be safest and simplest. Thoughts?

JDWoodcock commented 8 years ago

Stick with ratios- most people dont understand %s.

Robinlovelace commented 8 years ago

I disagree - I think more people understand % than "ratio", especially as the raw numbers are also provided, minimising the probability of mis-interpretation. One to discuss I think.

mvl22 commented 8 years ago

I think 'cyclists' is not the right word.

"Bicycle user" or "Cycle user" is what we use nowadays in campaigning in Cambridge. Don't know if that's what you meant.