single_view_embedding_for_sve: Single View Embedding (updated version)

In luke-a-rogers/pbsedm: An R package to implement some of the methods of empirical dynamic modelling

Single View Embedding (updated version)

Description

Andy taken Luke's original and adapting it. Cumbersome function name for now, but being consistent with other new function names. Call various functions to do a single view embedding for a given set of lags as per Hao and Sugihara.

Usage

single_view_embedding_for_sve(data, response, lags, ...)

Arguments

data	[matrix()] or [data.frame()] with named [numeric()] columns
response	[character()] column name of the response variable in data
lags	[list()] of a named vector of lags for each explanatory variable.
...	currently just 'max_allowed_correlation' to pass onto 'state_space_reconstruction_for_sve()'
metric	[character()]

Value

[tibble()] with columns 'response', 'response_predicted' (so both in original absolute numbers, and then 'response_s' and 'response_s_predicted' both in scaled (first differenced and then scaled) co-ordinates, so we can then calculate both types of correlation coefficient (or whatever else) to rank each individual single view embedding. OR a single NA if the lagged variables are highly correlated (see 'state_space_reconstruction_for_sve()'.

Author(s)

Andrew M. Edwards and Luke A. Rogers

Examples

## Not run: 
h_simulated <- 0.1095 + sample(1:180) * 0.001 # has mean of 0.2
simulated_4 <- EDMsimulate::salmon_sim(h = h_simulated)
res <- single_view_embedding_for_sve(data = simulated_4,
                             response = "R_t",
                             lags = create_subset_lags(list(R_t = 0:4,
                                                            S_t = 0:8
                                                            ))[[16000]]) # picking a specific subset of
                                                                         # potential lags
res %>% as.data.frame()
# Shows that can have R_t_predicted bigger than any original R_t, e.g. line
#  73, because R_t_s[72] was the largest possible, and previous R_t was not
#  very small. This may change with different seeds, as hadn't set, but idea
#  should hold.

## End(Not run)

luke-a-rogers/pbsedm documentation built on June 3, 2024, 5:20 a.m.