Use od_disaggregate() in od_jitter()

[Robinlovelace]

With #38 od_disaggregate() is more efficient, and deserves to be used in od_jitter().

[Robinlovelace]

They seem to have similar performance on a small dataset:

library(od)
od = od_data_df
z = od_data_zones_min
dlr = od_jitter(od, z) # desire_lines_random
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
#> Linking to GEOS 3.9.1, GDAL 3.3.2, PROJ 7.2.1
dlr2 = od_disaggregate(od, z, max_per_od = 10000, keep_ids = FALSE)
#> Creating origin and destination points at random locations

desire_lines = od_to_sf(od, z)
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
plot(z$geometry)
plot(dlr, add = TRUE, lwd = 3)
#> Warning in plot.sf(dlr, add = TRUE, lwd = 3): ignoring all but the first
#> attribute
plot(dlr2, add = TRUE, lwd = 3)
#> Warning in plot.sf(dlr2, add = TRUE, lwd = 3): ignoring all but the first
#> attribute
plot(desire_lines, add = TRUE, lwd = 5)
#> Warning in plot.sf(desire_lines, add = TRUE, lwd = 5): ignoring all but the
#> first attribute

dim(dlr)
#> [1]  7 11
dim(dlr2)
#> [1]  7 11
waldo::compare(sf::st_drop_geometry(dlr), sf::st_drop_geometry(dlr2)) # minor diffs
#> `names(old)[1:5]`: "geo_code1" "geo_code2" "all" "train" "bus"
#> `names(new)[1:5]`: "o"         "d"         "all" "train" "bus"
#> 
#> `old$geo_code1` is a character vector ('E02002384', 'E02002404', 'E02006875', 'E02006876', 'E02006861', ...)
#> `new$geo_code1` is absent
#> 
#> `old$geo_code2` is a character vector ('E02006875', 'E02006875', 'E02006875', 'E02006875', 'E02002392', ...)
#> `new$geo_code2` is absent
#> 
#> `old$o` is absent
#> `new$o` is a character vector ('1', '5', '13', '14', '6', ...)
#> 
#> `old$d` is absent
#> `new$d` is a character vector ('10', '7', '9', '8', '3', ...)
dlr$all == dlr$all
#> [1] TRUE TRUE TRUE TRUE TRUE TRUE TRUE

bench::mark(check = FALSE, 
            od_jitter = od_jitter(od, z),
            od_disaggregate = od_disaggregate(od, z, max_per_od = 10000, keep_ids = FALSE)
            )
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
#> Creating origin and destination points at random locations
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
#> 0 origins with no match in zone ids
#> 0 destinations with no match in zone ids
#>  points not in od data removed.
#> Creating origin and destination points at random locations
#> Creating origin and destination points at random locations
#> Creating origin and destination points at random locations
#> Warning: Some expressions had a GC in every iteration; so filtering is disabled.
#> # A tibble: 2 × 6
#>   expression           min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>      <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 od_jitter          190ms    190ms      5.13    6.26MB     5.13
#> 2 od_disaggregate    155ms    161ms      5.87    4.69MB     5.87

^{Created on 2021-12-05 by the reprex package (v2.0.1)}