In ecodata1/hera: Building Environmental Models Together

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  echo = FALSE
)
library(hera)
library(dplyr)
library(purrr)
library(tidyr)
library(readr)
library(magrittr)
library(testthat)

Welcome

This helps to create predictor variables for regulatory assessment.

Data Sources

Survey data

survey_inv <- read_csv("inst/extdat/predictor_raw_data/survey-inv.csv")
survey_inv <- survey_inv %>% filter(Determinand %in%
  c(
    "River Width (m)",
    "Mean Depth (cm)",
    "% Boulders/Cobbles",
    "% Pebbles/Gravel",
    "% Sand",
    "% Silt/Clay"
  ))

survey_inv <- survey_inv %>% mutate(Result = as.numeric(sub("<", "", Result)))
summary <- survey_inv %>%
  group_by(Determinand, `Location Code`) %>%
  summarise(
    mean_result = mean(Result, na.rm = TRUE)
  )

summary <- summary %>%
  pivot_wider(
    names_from = Determinand,
    values_from = mean_result
  )

predictors <- summary %>% rename(
  location_id = `Location Code`,
  boulders_cobbles = `% Boulders/Cobbles`,
  pebbles_gravel = `% Pebbles/Gravel`,
  silt_clay = `% Silt/Clay`,
  sand = `% Sand`,
  river_width = `River Width (m)`,
  mean_depth = `Mean Depth (cm)`
)

GIS and discharge data

sampling_points <- read_csv("inst/extdat/predictor_raw_data/sampling-points.csv")

sampling_points <- sampling_points %>% mutate(`S Ecology Alt Site` = ifelse(is.na(`S Ecology Alt Site`), Name, `S Ecology Alt Site`))

sampling_points <- sampling_points %>% mutate(`S Ecology Alt Site` = ifelse(`S Ecology Alt Site` == 0, Name, `S Ecology Alt Site`))

sampling_points <- sampling_points %>%
  filter(complete.cases(sampling_points))

sampling_points <- sampling_points %>%
  rename(
    location_id = "Name",
    grid_reference = "S Grid Reference",
    altitude = "S Altitude",
    slope = "S Slope",
    discharge_category = "S Discharge Cat",
    dist_from_source = "S Dist From Source",
    chemistry_site = "S Ecology Alt Site"
  )

Join survey and map based predictors

predictors <- full_join(sampling_points,
  predictors,
  by = c("location_id" = "location_id")
)

predictors <- predictors %>%
  mutate(total = silt_clay +
    sand +
    boulders_cobbles +
    pebbles_gravel)

Alkalinity

alkalinity <- read_csv("inst/extdat/predictor_raw_data/alkalinity.csv")

alk_summary <- alkalinity %>%
  group_by(`Location Code`) %>%
  summarise(
    alkalinity = mean(Result, na.rm = TRUE),
    alk_samples = n()
  )

alk_summary <- alk_summary %>%
  mutate(alkalinity = ifelse(alkalinity < 0, 0, alkalinity))

alk_summary <- alk_summary %>%
  rename(location_id = "Location Code")

Join alkalinity

predictors <- predictors %>%
  mutate(chemistry_site = ifelse(is.na(chemistry_site),
                                 location_id,
                                 chemistry_site))

predictors <- full_join(predictors,
  alk_summary,
  by = c("chemistry_site" = "location_id")
)

predictors <- predictors %>%
  filter(!is.na(location_id))

Save

write.csv(predictors, file = "predictors.csv", row.names = FALSE)

ecodata1/hera documentation built on July 17, 2025, 12:19 a.m.