CR request

[amd112]

Particular issues I'd like extra eyes on:

• 01_get_isochrones.R: I'm an SP user that's transitioning to SF so all my spatial work would be nice to look at with closer eyes. In particular the way I'm getting the population that has or doesn't have access to a pool is a little convoluted - I'd love for the logic to be checked for lines 103-132 to make sure your understanding of what's happening aligns. Another note for this file is that I added driving isochrones yesterday but haven't done anything with them yet, we were thinking it might be an additional useful measure.
• 02_map.R: I've been changing stuff in councildown in ways that interact with leaflet, I think as long as councildown is loaded last but something to be aware of. Make sure you have the most recent version!
• 03_summary_stat_map.R: These are static maps for the committee report, so would need a very quick turnaround on any comments here.
• 04_web_summary_stats.R: Pulls up the numbers referenced in the webpage, and creates a few barcharts for the webpage.

[romartinez-nycc]

01

Should we also set the CRS here?

• https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L65
• https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L75

Your code here works just as well for joining overlapping areas with st_interesect(),

https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L80

• Typically st_interesect() is used used as shown below & may be more concise:

  overlaps_area <- no_use %>%
  st_join(interest_area %>% st_as_sf(), st_intersects) 

  Getting error msg when rbinding the driving ischrones

  Error in bind_rows():! Argument 1 must be a data frame or a named atomic vector.

• https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L93

FIXED! #Getting error msg when pulling census pop data

Error in UseMethod("gather") : no applicable method for 'gather' applied to an object of class "character"

• https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L103

02

[amd112]

1) They are both natively ESPG 4326!

2) updated to overlaps_area <- no_use %>% st_join(interest_area, left = F) which passes the identical() check.

3) I haven't actually used any of the drive code yet so I think we should come back to this once that section is either implemented or removed [e: it was referencing pools_drive_zone_15min instead of land_drive_zone_15min - fixed]

4) what did you fix? what code did you change?

[romartinez-nycc]

FIXED! It wasn't reading my census API key, leading to a gather error, I didn't edit any of your code, only on my end I fixed the API key setup

Getting error msg when pulling census pop data

Error in UseMethod("gather") : no applicable method for 'gather' applied to an object of class "character"

• https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L103

We do now have estimated population distributions using bbls - which we found to be not far off from tracts & blocks

https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L119

Your logic/method makes sense for getting % or people within the pool zone

I get the same 68% number when summing up the blocks to get the 'near a pool percent'. The only issue on the website 68% is written as being not near a pool...

mutate(total_area = as.numeric(st_area(.)), 
area_within_pool_zone = replace_na(area_within_pool_zone, 0), 
perc_area = (total_area - area_within_pool_zone)/total_area,

     # get rid of tiny measurement errors like -0.000000001
     perc_area = ifelse(perc_area < 0, 0, perc_area), 

     # scale pop by (assumed) % not in zone
     pop = value * perc_area, 
     tot_pop = value) %>%

select(-total_area, -area_within_pool_zone, -value)

[romartinez-nycc]

1. They are both natively ESPG 4326!
2. updated to overlaps_area <- no_use %>% st_join(interest_area, left = F) which passes the identical() check.
3. I haven't actually used any of the drive code yet so I think we should come back to this once that section is either implemented or removed [e: it was referencing pools_drive_zone_15min instead of land_drive_zone_15min - fixed]
4. what did you fix? what code did you change?

Similar to #2 above, here it can be updated to use st_join() as well

https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L144 https://github.com/NewYorkCityCouncil/pool_access/blob/ba2bc980c65bcdcde64c0f8cf37bd31657a12507/code/01_get_isochrones.R#L147