diff --git a/DESCRIPTION b/DESCRIPTION
index 68c84e1..020669a 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -18,7 +18,7 @@ License: GPL-3
 Encoding: UTF-8
 LazyData: true
 Roxygen: list(markdown = TRUE)
-RoxygenNote: 7.1.2
+RoxygenNote: 7.3.2
 URL: https://github.com/gpw13/augury
 BugReports: https://github.com/gpw13/augury/issues
 Depends: 
diff --git a/R/predict_lme4_wrappers.R b/R/predict_lme4_wrappers.R
index 56ad7ce..7ee2792 100644
--- a/R/predict_lme4_wrappers.R
+++ b/R/predict_lme4_wrappers.R
@@ -7,6 +7,7 @@
 #' [predict_lme4()].
 #'
 #' @inherit predict_general_mdl params return
+#' @param REML Flag passed directly to lme4::lmer(). From `lme4` docs: "logical scalar - Should the estimates be chosen to optimize the REML criterion (as opposed to the log-likelihood)?"
 #'
 #' @export
 predict_lmer <- function(df,
@@ -36,7 +37,8 @@ predict_lmer <- function(df,
                          replace_obs = c("missing", "all", "none"),
                          error_correct = FALSE,
                          error_correct_cols = NULL,
-                         shift_trend = FALSE) {
+                         shift_trend = FALSE,
+                         REML = TRUE) {
   predict_lme4(df = df,
                model = lme4::lmer,
                formula = formula,
@@ -65,7 +67,8 @@ predict_lmer <- function(df,
                replace_obs = replace_obs,
                error_correct = error_correct,
                error_correct_cols = error_correct_cols,
-               shift_trend = shift_trend)
+               shift_trend = shift_trend,
+               REML = REML)
 }
 
 #' Use a generalized linear mixed-effects model to infill and project data
diff --git a/man/fit_aarr_model.Rd b/man/fit_aarr_model.Rd
index 27deb1d..ee4d93e 100644
--- a/man/fit_aarr_model.Rd
+++ b/man/fit_aarr_model.Rd
@@ -58,12 +58,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{pred_col}{Column name to store predicted value.}
 }
diff --git a/man/fit_forecast_average_model.Rd b/man/fit_forecast_average_model.Rd
index 71f88bf..30a910c 100644
--- a/man/fit_forecast_average_model.Rd
+++ b/man/fit_forecast_average_model.Rd
@@ -27,7 +27,7 @@ fit_forecast_average_model(
 \item{df}{Data frame of model data.}
 
 \item{forecast_function}{An R function that outputs a forecast object coming from the
-forecast package. You can directly pass \code{\link[forecast:forecast]{forecast::forecast()}} to the
+forecast package. You can directly pass \code{\link[forecast:reexports]{forecast::forecast()}} to the
 function, or you can pass other wrappers to it such as \code{\link[forecast:ses]{forecast::holt()}} or
 \code{\link[forecast:ses]{forecast::ses()}}.}
 
diff --git a/man/fit_forecast_model.Rd b/man/fit_forecast_model.Rd
index 5da86c3..60cb8f4 100644
--- a/man/fit_forecast_model.Rd
+++ b/man/fit_forecast_model.Rd
@@ -26,7 +26,7 @@ fit_forecast_model(
 \item{df}{Data frame of model data.}
 
 \item{forecast_function}{An R function that outputs a forecast object coming from the
-forecast package. You can directly pass \code{\link[forecast:forecast]{forecast::forecast()}} to the
+forecast package. You can directly pass \code{\link[forecast:reexports]{forecast::forecast()}} to the
 function, or you can pass other wrappers to it such as \code{\link[forecast:ses]{forecast::holt()}} or
 \code{\link[forecast:ses]{forecast::ses()}}.}
 
@@ -54,12 +54,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/fit_general_model.Rd b/man/fit_general_model.Rd
index 56319f9..f1db389 100644
--- a/man/fit_general_model.Rd
+++ b/man/fit_general_model.Rd
@@ -60,12 +60,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/fit_inla_average_model.Rd b/man/fit_inla_average_model.Rd
index ecbe686..767d2de 100644
--- a/man/fit_inla_average_model.Rd
+++ b/man/fit_inla_average_model.Rd
@@ -44,7 +44,7 @@ of the fitted values returned for use in the infilling and predictions. Addition
 arguments can be passed in the \code{control.predictor} list, but must always include
 \code{compute = TRUE}. See \code{\link[INLA:control.predictor]{INLA::control.predictor()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{formula_vars}{Variables included in the model formula, generated by
 \code{all.vars(formula)}.}
@@ -84,7 +84,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
diff --git a/man/fit_inla_model.Rd b/man/fit_inla_model.Rd
index 4815807..ef10a83 100644
--- a/man/fit_inla_model.Rd
+++ b/man/fit_inla_model.Rd
@@ -37,7 +37,7 @@ of the fitted values returned for use in the infilling and predictions. Addition
 arguments can be passed in the \code{control.predictor} list, but must always include
 \code{compute = TRUE}. See \code{\link[INLA:control.predictor]{INLA::control.predictor()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{formula_vars}{Variables included in the model formula, generated by
 \code{all.vars(formula)}.}
@@ -61,12 +61,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -90,7 +92,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
diff --git a/man/fit_lme4_model.Rd b/man/fit_lme4_model.Rd
index a08c796..f8ea773 100644
--- a/man/fit_lme4_model.Rd
+++ b/man/fit_lme4_model.Rd
@@ -59,12 +59,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/forecast_series.Rd b/man/forecast_series.Rd
index 627cb8b..b30e8bb 100644
--- a/man/forecast_series.Rd
+++ b/man/forecast_series.Rd
@@ -10,7 +10,7 @@ forecast_series(x, forecast_function, ...)
 \item{x}{Series to forecast, coming from \code{trim_series()}}
 
 \item{forecast_function}{An R function that outputs a forecast object coming from the
-forecast package. You can directly pass \code{\link[forecast:forecast]{forecast::forecast()}} to the
+forecast package. You can directly pass \code{\link[forecast:reexports]{forecast::forecast()}} to the
 function, or you can pass other wrappers to it such as \code{\link[forecast:ses]{forecast::holt()}} or
 \code{\link[forecast:ses]{forecast::ses()}}.}
 
diff --git a/man/map_model_behavior.Rd b/man/map_model_behavior.Rd
index b439167..19d6669 100644
--- a/man/map_model_behavior.Rd
+++ b/man/map_model_behavior.Rd
@@ -17,12 +17,12 @@ of the fitted values returned for use in the infilling and predictions. Addition
 arguments can be passed in the \code{control.predictor} list, but must always include
 \code{compute = TRUE}. See \code{\link[INLA:control.predictor]{INLA::control.predictor()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 }
 \value{
 A data frame.
 }
 \description{
-\code{map_model_behavior()} runs \code{\link[INLA:inla]{INLA::inla()}} until its successful. This is
+\code{map_model_behavior()} runs \code{\link[INLA:INLA]{INLA::inla()}} until its successful. This is
 important on  linux machines where INLA can produce unstable results.
 }
diff --git a/man/merge_average_df.Rd b/man/merge_average_df.Rd
index c9273f3..c5c482e 100644
--- a/man/merge_average_df.Rd
+++ b/man/merge_average_df.Rd
@@ -39,12 +39,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/merge_prediction.Rd b/man/merge_prediction.Rd
index a04b6c6..7809113 100644
--- a/man/merge_prediction.Rd
+++ b/man/merge_prediction.Rd
@@ -41,12 +41,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_aarr.Rd b/man/predict_aarr.Rd
index 3a44ac9..0ec4a89 100644
--- a/man/predict_aarr.Rd
+++ b/man/predict_aarr.Rd
@@ -90,12 +90,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/predict_average.Rd b/man/predict_average.Rd
index ed20a63..0edbbea 100644
--- a/man/predict_average.Rd
+++ b/man/predict_average.Rd
@@ -73,12 +73,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_average_fn.Rd b/man/predict_average_fn.Rd
index 1f471ef..2ec5d48 100644
--- a/man/predict_average_fn.Rd
+++ b/man/predict_average_fn.Rd
@@ -50,12 +50,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{pred_col}{Column name to store predicted value.}
 
diff --git a/man/predict_forecast.Rd b/man/predict_forecast.Rd
index 7eb9be1..b8a6fe1 100644
--- a/man/predict_forecast.Rd
+++ b/man/predict_forecast.Rd
@@ -39,7 +39,7 @@ predict_forecast(
 \item{df}{Data frame of model data.}
 
 \item{forecast_function}{An R function that outputs a forecast object coming from the
-forecast package. You can directly pass \code{\link[forecast:forecast]{forecast::forecast()}} to the
+forecast package. You can directly pass \code{\link[forecast:reexports]{forecast::forecast()}} to the
 function, or you can pass other wrappers to it such as \code{\link[forecast:ses]{forecast::holt()}} or
 \code{\link[forecast:ses]{forecast::ses()}}.}
 
@@ -94,12 +94,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
@@ -165,7 +167,7 @@ Depending on the value passed to \code{ret}, either a data frame with
 predicted data, a vector of errors from \code{\link[=model_error]{model_error()}}, a fitted model, or a list with all 3.
 }
 \description{
-\code{predict_forecast()} uses the forecast package's \code{\link[forecast:forecast]{forecast::forecast()}} methods
+\code{predict_forecast()} uses the forecast package's \code{\link[forecast:reexports]{forecast::forecast()}} methods
 to generate predictions on time series data. These use the longest contiguous
 observed values to forecast out a certain number periods. This function
 automatically detects the latest observed values and the number of missing
diff --git a/man/predict_forecast_avg_trend.Rd b/man/predict_forecast_avg_trend.Rd
index 976a240..f327658 100644
--- a/man/predict_forecast_avg_trend.Rd
+++ b/man/predict_forecast_avg_trend.Rd
@@ -41,7 +41,7 @@ predict_forecast_avg_trend(
 \item{df}{Data frame of model data.}
 
 \item{forecast_function}{An R function that outputs a forecast object coming from the
-forecast package. You can directly pass \code{\link[forecast:forecast]{forecast::forecast()}} to the
+forecast package. You can directly pass \code{\link[forecast:reexports]{forecast::forecast()}} to the
 function, or you can pass other wrappers to it such as \code{\link[forecast:ses]{forecast::holt()}} or
 \code{\link[forecast:ses]{forecast::ses()}}.}
 
@@ -102,12 +102,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
@@ -175,7 +177,7 @@ predicted data, a vector of errors from \code{\link[=model_error]{model_error()}
 \description{
 \code{predict_forecast_avg_trend()} uses time series forecasting methods to fit a model
 to groups within the data, and then bring that fitted prediction back to the
-original data. The function uses \code{\link[forecast:forecast]{forecast::forecast()}}, and full details and
+original data. The function uses \code{\link[forecast:reexports]{forecast::forecast()}}, and full details and
 explanation of arguments that it can accept is available there. The augury
 implementation details are available at \code{\link[=predict_forecast]{predict_forecast()}}.
 The function also allows for inputting of data type and source information
diff --git a/man/predict_general_mdl.Rd b/man/predict_general_mdl.Rd
index 753519e..8d5a27f 100644
--- a/man/predict_general_mdl.Rd
+++ b/man/predict_general_mdl.Rd
@@ -96,12 +96,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_glm.Rd b/man/predict_glm.Rd
index 052a9e2..d37b9bb 100644
--- a/man/predict_glm.Rd
+++ b/man/predict_glm.Rd
@@ -79,12 +79,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_glmer.Rd b/man/predict_glmer.Rd
index 39d4477..c903686 100644
--- a/man/predict_glmer.Rd
+++ b/man/predict_glmer.Rd
@@ -79,12 +79,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_glmer_avg_trend.Rd b/man/predict_glmer_avg_trend.Rd
index 9f649ef..e2c632a 100644
--- a/man/predict_glmer_avg_trend.Rd
+++ b/man/predict_glmer_avg_trend.Rd
@@ -98,12 +98,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_holt.Rd b/man/predict_holt.Rd
index 625ba14..4247ace 100644
--- a/man/predict_holt.Rd
+++ b/man/predict_holt.Rd
@@ -77,12 +77,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/predict_holt_avg_trend.Rd b/man/predict_holt_avg_trend.Rd
index a43ad79..cd94d51 100644
--- a/man/predict_holt_avg_trend.Rd
+++ b/man/predict_holt_avg_trend.Rd
@@ -96,12 +96,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/predict_inla.Rd b/man/predict_inla.Rd
index e0af51b..2e439cb 100644
--- a/man/predict_inla.Rd
+++ b/man/predict_inla.Rd
@@ -49,7 +49,7 @@ of the fitted values returned for use in the infilling and predictions. Addition
 arguments can be passed in the \code{control.predictor} list, but must always include
 \code{compute = TRUE}. See \code{\link[INLA:control.predictor]{INLA::control.predictor()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
 to returning a data frame, but can return a vector of model error, the
@@ -97,12 +97,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -136,7 +138,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{type_col}{Column name specifying data type.}
@@ -185,7 +187,7 @@ predicted data, a vector of errors from \code{\link[=model_error]{model_error()}
 \description{
 \code{predict_inla()} uses a Integrated Nested Laplace approximation to fit a model
 and use that model to infill and project the dependent variable. The function
-uses \code{\link[INLA:inla]{INLA::inla()}} to perform the model fitting and prediction, and full
+uses \code{\link[INLA:INLA]{INLA::inla()}} to perform the model fitting and prediction, and full
 details and explanation of arguments that it can accept is available on that page.
 The function also allows for inputting of data type and source information
 directly into the data frame if the \code{type_col} and \code{source_col} are specified
diff --git a/man/predict_inla_avg_trend.Rd b/man/predict_inla_avg_trend.Rd
index 452142b..9cfb0e2 100644
--- a/man/predict_inla_avg_trend.Rd
+++ b/man/predict_inla_avg_trend.Rd
@@ -61,7 +61,7 @@ of the fitted values returned for use in the infilling and predictions. Addition
 arguments can be passed in the \code{control.predictor} list, but must always include
 \code{compute = TRUE}. See \code{\link[INLA:control.predictor]{INLA::control.predictor()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
 to returning a data frame, but can return a vector of model error, the
@@ -105,12 +105,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -144,7 +146,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{type_col}{Column name specifying data type.}
@@ -193,7 +195,7 @@ predicted data, a vector of errors from \code{\link[=model_error]{model_error()}
 \description{
 \code{predict_inla_avg_trend()} uses a Integrated Nested Laplace approximation to fit a model
 to groups within the data, and then bring that fitted prediction back to the
-original data. The function uses \code{\link[INLA:inla]{INLA::inla()}} to perform the model fitting
+original data. The function uses \code{\link[INLA:INLA]{INLA::inla()}} to perform the model fitting
 and prediction, and full details and explanation of arguments that it can accept is available on that page.
 The function also allows for inputting of data type and source information
 directly into the data frame if the \code{type_col} and \code{source_col} are specified
diff --git a/man/predict_inla_data.Rd b/man/predict_inla_data.Rd
index cd956ec..a9bd6cc 100644
--- a/man/predict_inla_data.Rd
+++ b/man/predict_inla_data.Rd
@@ -9,7 +9,7 @@ predict_inla_data(df, model, pred_col, pred_upper_col, pred_lower_col)
 \arguments{
 \item{df}{Data frame of model data.}
 
-\item{model}{INLA model object returned by \code{\link[INLA:inla]{INLA::inla()}} which contains
+\item{model}{INLA model object returned by \code{\link[INLA:INLA]{INLA::inla()}} which contains
 \code{summary.fitted.values}.}
 
 \item{pred_col}{Column name to store predicted value.}
@@ -26,6 +26,6 @@ values for the lower bound.}
 A data frame.
 }
 \description{
-\code{predict_inla_data()} generates a prediction vector from an \code{\link[INLA:inla]{INLA::inla()}}
+\code{predict_inla_data()} generates a prediction vector from an \code{\link[INLA:INLA]{INLA::inla()}}
 output object, putting this prediction back into the data frame.
 }
diff --git a/man/predict_inla_me.Rd b/man/predict_inla_me.Rd
index cc6238a..23a78fb 100644
--- a/man/predict_inla_me.Rd
+++ b/man/predict_inla_me.Rd
@@ -7,8 +7,8 @@
 predict_inla_me(
   df,
   model = "rw2",
- 
-    formula = stats::as.formula(sprintf("value ~ sdi_scaled + e0_scaled + f(region, model = 'iid') + f(year, model = '\%s')",
+  formula =
+    stats::as.formula(sprintf("value ~ sdi_scaled + e0_scaled + f(region, model = 'iid') + f(year, model = '\%s')",
     model)),
   control.predictor = list(compute = TRUE),
   family = "gaussian",
@@ -62,7 +62,7 @@ detailed docs for individual families.}
 
 \item{control.inla}{See \code{\link[INLA:control.inla]{INLA::control.inla()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
 to returning a data frame, but can return a vector of model error, the
@@ -101,12 +101,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -140,7 +142,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{type_col}{Column name specifying data type.}
diff --git a/man/predict_inla_ts.Rd b/man/predict_inla_ts.Rd
index e5124fb..bebaa12 100644
--- a/man/predict_inla_ts.Rd
+++ b/man/predict_inla_ts.Rd
@@ -60,7 +60,7 @@ arguments can be passed in the \code{control.predictor} list, but must always in
 
 \item{control.inla}{See \code{\link[INLA:control.inla]{INLA::control.inla()}} for details.}
 
-\item{...}{Additional arguments passed to \code{\link[INLA:inla]{INLA::inla()}}.}
+\item{...}{Additional arguments passed to \code{\link[INLA:INLA]{INLA::inla()}}.}
 
 \item{ret}{Character vector specifying what values the function returns. Defaults
 to returning a data frame, but can return a vector of model error, the
@@ -99,12 +99,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -138,7 +140,7 @@ values from the dataset prior to applying the model. By default, only
 observations with missing predictors are removed, although it can also remove
 rows only if they have missing dependent or independent variables, or no
 filtering at all. Model prediction and fitting are done in one pass with
-\code{\link[INLA:inla]{INLA::inla()}}, so there will be no predictions if observations with
+\code{\link[INLA:INLA]{INLA::inla()}}, so there will be no predictions if observations with
 missing dependent variables are removed.}
 
 \item{type_col}{Column name specifying data type.}
diff --git a/man/predict_lm.Rd b/man/predict_lm.Rd
index b9af060..6a2cf4e 100644
--- a/man/predict_lm.Rd
+++ b/man/predict_lm.Rd
@@ -79,12 +79,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_lme4.Rd b/man/predict_lme4.Rd
index 9612da0..565fe32 100644
--- a/man/predict_lme4.Rd
+++ b/man/predict_lme4.Rd
@@ -95,12 +95,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_lme4_avg_trend.Rd b/man/predict_lme4_avg_trend.Rd
index d2ab9f4..76f06d1 100644
--- a/man/predict_lme4_avg_trend.Rd
+++ b/man/predict_lme4_avg_trend.Rd
@@ -103,12 +103,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_lmer.Rd b/man/predict_lmer.Rd
index 7cb5aa0..a4b2037 100644
--- a/man/predict_lmer.Rd
+++ b/man/predict_lmer.Rd
@@ -32,7 +32,8 @@ predict_lmer(
   replace_obs = c("missing", "all", "none"),
   error_correct = FALSE,
   error_correct_cols = NULL,
-  shift_trend = FALSE
+  shift_trend = FALSE,
+  REML = TRUE
 )
 }
 \arguments{
@@ -79,12 +80,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
@@ -157,6 +160,8 @@ error correction to the predicted values.}
 \item{shift_trend}{Logical value specifying whether or not to shift predictions
 so that the trend matches up to the last observation. If \code{error_correct} and
 \code{shift_trend} are both \code{TRUE}, \code{shift_trend} takes precedence.}
+
+\item{REML}{Flag passed directly to lme4::lmer(). From \code{lme4} docs: "logical scalar - Should the estimates be chosen to optimize the REML criterion (as opposed to the log-likelihood)?"}
 }
 \value{
 Depending on the value passed to \code{ret}, either a data frame with
diff --git a/man/predict_lmer_avg_trend.Rd b/man/predict_lmer_avg_trend.Rd
index ad1d607..7144846 100644
--- a/man/predict_lmer_avg_trend.Rd
+++ b/man/predict_lmer_avg_trend.Rd
@@ -98,12 +98,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_nlmer.Rd b/man/predict_nlmer.Rd
index 41bdb52..ae361b5 100644
--- a/man/predict_nlmer.Rd
+++ b/man/predict_nlmer.Rd
@@ -79,12 +79,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_nlmer_avg_trend.Rd b/man/predict_nlmer_avg_trend.Rd
index 011004e..4d382af 100644
--- a/man/predict_nlmer_avg_trend.Rd
+++ b/man/predict_nlmer_avg_trend.Rd
@@ -98,12 +98,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to \code{\link[dplyr:arrange]{dplyr::arrange()}} the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_ses.Rd b/man/predict_ses.Rd
index e872872..a9641a1 100644
--- a/man/predict_ses.Rd
+++ b/man/predict_ses.Rd
@@ -77,12 +77,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/predict_ses_avg_trend.Rd b/man/predict_ses_avg_trend.Rd
index 5792dcd..4273e50 100644
--- a/man/predict_ses_avg_trend.Rd
+++ b/man/predict_ses_avg_trend.Rd
@@ -96,12 +96,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name of column to arrange data by in \code{dplyr::arrange()},
 prior to filtering for latest contiguous time series and producing the
diff --git a/man/predict_simple.Rd b/man/predict_simple.Rd
index 43bb850..ad411e4 100644
--- a/man/predict_simple.Rd
+++ b/man/predict_simple.Rd
@@ -70,12 +70,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{sort_col}{Column name(s) to use to dplyr::arrange() the data prior to
 supplying type and calculating mean absolute scaled error on data involving
diff --git a/man/predict_simple_fn.Rd b/man/predict_simple_fn.Rd
index 8228e10..e572c02 100644
--- a/man/predict_simple_fn.Rd
+++ b/man/predict_simple_fn.Rd
@@ -46,12 +46,14 @@ replace observations with predicted values. This is done in
 conjunction with \code{group_col}. So, if \code{group_col = "iso3"} and
 \code{obs_filter = ">= 5"}, then for this model, predictions will only be used
 for \code{iso3} vales that have 5 or more observations. Possible logical operators
-to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
+to use are \code{>}, \code{>=}, \code{<}, \code{<=}, \code{==}, and \code{!=}.
+
+\if{html}{\out{<div class="sourceCode">}}\preformatted{If `group_models = FALSE`, then `obs_filter` is only used to determine when
 predicted values replace observed values but **is not** used to restrict values
 from being used in model fitting. If `group_models = TRUE`, then a model
 is only fit for a group if they meet the `obs_filter` requirements. This provides
 speed benefits, particularly when running INLA time series using `predict_inla()`.
-}}
+}\if{html}{\out{</div>}}}
 
 \item{pred_col}{Column name to store predicted value.}