Define a non-linear relationship

Closure to be used in sim_distribution.

sim_nlf(
  formula = ~alpha - ((depth - mu)^2)/(2 * sigma^2),
  coeff = list(alpha = 0, mu = 200, sigma = 70)
)

Arguments

formula

Formula describing parametric relationships between data and coefficients. The data used in sim_distribution are grid coordinates expanded across ages and years (i.e., includes columns "x", "y", "depth", "cell", "division", "strat", "age", "year"). Values of the coefficients must be included in argument coeff as a named list.

coeff

Named list of coefficient values used in formula.

Value

Returns a function for use inside sim_distribution.

Examples

## Make a grid and replicate data for 5 ages and 5 years
## (This is similar to what happens inside sim_distribution)
grid <- make_grid(shelf_width = 10)
grid_xy <- data.frame(grid)
i <- rep(seq(nrow(grid_xy)), times = 5)
a <- rep(1:5, each = nrow(grid_xy))
grid_xy <- grid_xy[i, ]
grid_xy$age <- a
i <- rep(seq(nrow(grid_xy)), times = 5)
y <- rep(1:5, each = nrow(grid_xy))
grid_xy <- grid_xy[i, ]
grid_xy$year <- y

## Now using sim_nlf, produce a function to apply to the expanded grid_xy data
## For this firs example, the depth effect is parabolic and the vertex is deeper by age
## (i.e., to impose ontogenetic deepening)
nlf <- sim_nlf(formula = ~ alpha - ((depth - mu + beta * age) ^ 2) / (2 * sigma ^ 2),
               coeff = list(alpha = 0, mu = 200, sigma = 70, beta = -70))
grid_xy$depth_effect <- nlf(grid_xy)

library(plotly)
#> Loading required package: ggplot2
#> 
#> Attaching package: ‘plotly’
#> The following object is masked from ‘package:ggplot2’:
#> 
#>     last_plot
#> The following object is masked from ‘package:stats’:
#> 
#>     filter
#> The following object is masked from ‘package:graphics’:
#> 
#>     layout
grid_xy %>%
  filter(year == 1) %>%
  plot_ly(x = ~depth, y = ~depth_effect, split = ~age) %>%
  add_lines()