Skip to contents

Carry each visit's minimal-network parent onto its full-network segment

Source: R/parent_join.R
knw_parent_join.Rd

link deploys a rule on its full network, whose segments are cut at fish records, crossings and habitat end points as well as at barriers and access gradients; its minimal network (default_minimal, link#415) is cut at the last two alone. A rule tuned on the minimal network transfers to the full one when each full segment carries its minimal parent's gradient (link#444). This joins that parent to the visits placed on the full network, so the tuning can read the parent's gradient there (GRADIENT=parent_gradient in scripts/0320, #120).

Usage

knw_parent_join(visits, parent)

Arguments

visits

data.frame of placed visits with site_visit, watershed_group_code, id_segment and segment_length_m (the full network's segment, as knw_link_place() returns it).

parent

data.frame with one row per full segment: id_segment, wsg, id_segment_minimal, gradient_minimal, length_minimal, overlap_m and straddle_m, as link's data-raw/logs/minimal_breaks_415/parent_<wsg>.csv.gz holds them.

Value

visits with parent_id_segment, parent_gradient, parent_length_m and parent_straddle_m, in the same rows and order.

Details

A full segment's parent is the minimal segment it overlaps most, and every visit on the full segment takes that parent's gradient. Where a full segment straddles a minimal break, straddle_m is the length that lies on another minimal segment: a visit on that stretch sits on the other segment on the minimal network, so there the two setups read different gradients. A placed visit whose segment has no parent row is an error. link numbers each group's segments from 1, so a parent table of another build covers the same ids: each visit's segment must be as long as its parent row's overlap_m + straddle_m (within 1 cm; a missing length fails), or the join stops as another network's table.

Examples

v <- data.frame(site_visit = c("a", "b"), watershed_group_code = "BULK", id_segment = c(10, 11),
                segment_length_m = c(120, 315))
p <- data.frame(id_segment = c(10, 11, 12), wsg = "BULK", id_segment_minimal = c(4, 4, 5),
                gradient_minimal = c(0.012, 0.012, 0.031), length_minimal = c(900, 900, 400),
                overlap_m = c(120, 300, 400), straddle_m = c(0, 15, 0))
knw_parent_join(v, p)
#>   site_visit watershed_group_code id_segment segment_length_m parent_id_segment
#> 1          a                 BULK         10              120                 4
#> 2          b                 BULK         11              315                 4
#>   parent_gradient parent_length_m parent_straddle_m
#> 1           0.012             900                 0
#> 2           0.012             900                15