OLS_modeling: functions related to modeling

In jstrunk001/RForInvt: Forest Inventory Data Compilation and FVS Growth Projections

functions related to modeling

Description

<Delete and Replace>

Usage

reg_model(
  dat_yx,
  reg_obj,
  rank = 1,
  rank_by = c("bic", "rsq", "rss", "adjr2", "cp"),
  debug = FALSE,
  ...
)

lm_boot(model, y = NA, data = NA, n_boot = 50)

many_boots(model, n_range, r_boots = 10, n_clus = 1, lm_boot, ...)

reg_multi(
  y_vars,
  data,
  form_y = c("y~.", "y~.*."),
  n_v_max = 3,
  n_best = 1,
  really_big = FALSE,
  n_clus = 1,
  ...
)

lm_multi(y_vars, data, form_y = c("y~.", "y~.*."), verbose = T, ...)

mod_multi(mods_list, data, ...)

pred_multi(
  mods_list,
  data,
  dat0,
  id_col,
  n_clus = 1,
  clus = NA,
  out_dir = NA,
  se_fit = FALSE,
  return = TRUE,
  fix_outliers = F,
  outlier_fun = function(y0, yhat, yhat_se) {
     large = yhat > max(y0) * 1.5
    
    small = yhat < 0
     yhat[small] = 0
     yhat[large] = quantile(y0, 0.99) * 1.5
   
     yhat_se[large] = max(yhat_se[!large & !small])
     yhat_se[small] =
    max(yhat_se[!large & !small])
     data.frame(pd = yhat, se_pd = yhat_se)
 }
)

multi_bs(mods_list, n_boot = 50, n_clus, lm_boot)

lm_summary(mods_list, data, resids = FALSE)

Arguments

dat_yx	tbd
reg_obj	tbd
rank	tbd
rank_by	tbd
debug	tbd
...	tbd pred_multi() parameters:
model	tbd
y	tbd
data	tbd
n_boot	tbd
n_range	tbd
r_boots	tbd
n_clus	tbd
lm_boot	tbd lm_summary() parameters:
y_vars	tbd
form_y	tbd
n_v_max	tbd
n_best	tbd
really_big	tbd
mods_list	list of models where each element is named with response name
dat0	tbd
id_col	tbd
clus	tbd
out_dir	tbd
se_fit	tbd
return	tbd
fix_outliers	tbd
outlier_fun	tbd multi_bs() parameters:
resids	tbd
mods	tbd

Details

<Delete and Replace>

Revision History

1.0	date and revisions..

Value

reg_model: lm_boot: many_boots: reg_multi: mod_multi: pred_multi: multi_bs: lm_summary:

Author(s)

Jacob Strunk <[e-mail]>

See Also

regsubsets

Examples

<Delete and Replace>

jstrunk001/RForInvt documentation built on April 17, 2025, 5:02 p.m.