Closed joeytalbot closed 4 years ago
Great idea.
I think Jobs5000EmpPTt
should be sufficient but yes, maybe a weighted mean with weights being proportional to the size. Example below:
devtools::install_github("cyipt/acton")
#> Skipping install of 'acton' from a github remote, the SHA1 (6c4468f7) has not changed since last install.
#> Use `force = TRUE` to force installation
library(acton)
jts = get_jts_data("jts0501", 2017)
#> This table's title is Travel time, destination and origin indicators for Employment centres by mode of travel, Lower Super Output Area (LSOA), England
#> These data files are available for that table code: jts0501-2014.csv
#> jts0501-2015.csv
#> jts0501-2016.csv
#> jts0501-2017.csv
#> jts0501-Metadata.csv
#> Reading in file https://github.com/cyipt/acton/releases/download/0.0.1/jts0501-2017.csv
#> Parsed with column specification:
#> cols(
#> .default = col_double(),
#> LSOA_code = col_character(),
#> Region = col_character(),
#> LA_Code = col_character(),
#> LA_Name = col_character()
#> )
#> See spec(...) for full column specifications.
summary(jts$Jobs100EmpPTt)
#> Min. 1st Qu. Median Mean 3rd Qu. Max.
#> 0.9689 4.6951 7.1631 9.2181 10.9838 120.0000
summary(jts$Jobs500EmpPTt)
#> Min. 1st Qu. Median Mean 3rd Qu. Max.
#> 0.8493 6.7811 10.8665 12.2315 15.3687 114.9730
summary(jts$Jobs5000EmpPTt)
#> Min. 1st Qu. Median Mean 3rd Qu. Max.
#> 1.061 19.886 27.698 32.575 39.338 120.000
jts$weighted_mean_time_to_employment_centre = apply(
X = jts[c("Jobs100EmpPTt", "Jobs500EmpPTt", "Jobs5000EmpPTt")],
MARGIN = 1,
FUN = weighted.mean,
w = c(100, 500, 5000)
)
summary(jts$weighted_mean_time_to_employment_centre)
#> Min. 1st Qu. Median Mean 3rd Qu. Max.
#> 1.727 18.786 25.886 30.341 36.462 118.464
Created on 2020-01-15 by the reprex package (v0.3.0)
Yes that's the kind of thing I was thinking of. Making the weights exactly proportional the to number of jobs is a great idea.
Even though it's the 'minimum number of jobs' I think, still worthwhile I think.
Could you make something like the h-index score for academics https://en.wikipedia.org/wiki/H-index. Aso a score would incorporate both number of jobs and travel time?
There are dozens of employment measures in the accessibility statistics. To make things clearer, we should be able to develop a simple weighted measure for each travel mode (walking/public transport, cycling and car).
On its own, I don't think
Jobs5000EmpPTt
is sufficient, but it will be better when combined with the stats on smaller employment centres (Jobs500EmpPTt
andJobs100EmpPTt
).