feature(plot_indices): create new plot of indices by fleet - fully fu…

…nctional and tested

Testing not performed on rda

NAMESPACE

@@ -3,6 +3,7 @@
 export(add_theme)
 export(exp_all_figs_tables)
 export(plot_biomass)
+export(plot_indices)
 export(plot_landings)
 export(plot_recruitment)
 export(plot_recruitment_deviations)

R/plot_indices.R

@@ -0,0 +1,157 @@
+#' Plot Index of Abundance
+#'
+#' @inheritParams plot_recruitment
+#' @param unit_label units for index of abundance/CPUE
+#'
+#' @return Plot the estimated indices as indicated from a standard assessment
+#' model output file.
+#' @export
+#'
+plot_indices <- function(
+    dat,
+    unit_label = NULL,
+    make_rda = TRUE,
+    rda_dir = NULL
+) {
+  # Set cpue unit label for plot
+  u_units <- glue::glue("Estimaed CPUE ({unit_label})")
+
+  # Load data
+  output <- dat |>
+    dplyr::filter(module_name == "INDEX_2" | module_name == "t.series")
+  # Check for U
+  if (any(unique(output$module_name=="INDEX_2"))) {
+    output <- output |>
+      dplyr::filter(grepl("expected_indices", label)) # grepl("input_indices", label) |
+  } else if (any(unique(output$module_name=="t.series"))) {
+    output <- output |>
+      dplyr::filter(grepl("cpue", label))
+  }
+  # Extract fleet names
+  fleet_names <- unique(as.character(output$fleet))
+  factors <- c("year", "fleet", "fleet_name", "age", "sex", "area", "seas", "season", "time", "era", "subseas", "subseason", "platoon", "platoo","growth_pattern", "gp")
+  # re-structure df for table
+  indices <- output |>
+    tidyr::pivot_wider(
+      # id_cols = c(year, uncertainty, uncertainty_label),
+      names_from = label,
+      values_from = estimate
+    ) |>
+    dplyr::select(year, fleet, unique(output$label), uncertainty, uncertainty_label) # |>
+
+  # na.omit()
+  # check if uncertainty is a measure in the df
+  if(all(is.na(indices$uncertainty))){
+    indices <- indices |>
+      dplyr::select(-c(uncertainty_label, uncertainty))
+  } else {
+    uncertainty_col <- paste("uncertainty_", unique(indices$uncertainty_label), sep = "")
+    colnames(indices) <- stringr::str_replace(colnames(indices), "^uncertainty$", uncertainty_col)
+    indices <- dplyr::select(indices, -uncertainty_label)
+  }
+
+  # Check if observed/inital values are in the df
+  if (any(grepl("observed", colnames(indices)))) {
+    indices <- indices |>
+      dplyr::select(-colnames(indices)[grep(c("observed"), colnames(indices))])
+  }
+
+  # rename columns to remove cpue/effort
+  if(any(grep("_indices", colnames(indices)))){
+    colnames(indices) <- stringr::str_replace_all(colnames(indices), "_indices", "")
+  } else {
+    colnames(indices) <- stringr::str_replace_all(colnames(indices), "cpue_", "")
+  }
+
+  # Check for which column is U and filter out na values
+  if (any(grep("predicted", colnames(indices)))) {
+    indices <- indices |>
+      dplyr::filter(!is.na(predicted)) # |>
+      # dplyr::rename(estimated = predicted)
+  }
+  if (any(grep("expected", colnames(indices)))) {
+    indices <- indices |>
+      dplyr::filter(!is.na(expected)) # |>
+      # dplyr::rename(estimated = predicted)
+  }
+
+  # Check for correct number of columns in dataframe
+  if (4 < ncol(indices)| ncol(indices) > 4) stop("Incorrect number of columns. Additional factor present.")
+  # Check for error type to present to user
+  # Double check this warning is correct
+  # if (grep("cv", colnames(indices)[4])) warning("Confidence Intervals were calculated using cv rather than se. Please use caution interpreting of the output plot.")
+  # Create condition for error type
+  if (grepl("cv", colnames(indices)[4])) {
+    err_val <- TRUE # indicates that the error value is cv
+  } else {
+    err_val <- FALSE # indicated that the error value is se
+  }
+
+  # Rename column names for easier plotting
+  # This will break if columns are not in the correct order - add check?
+  colnames(indices) <- c("year", "fleet", "estimate", "uncertainty")
+
+  # Final data set for plotting
+  indices2 <- indices |>
+    dplyr::mutate(
+      estimate_lower = dplyr::case_when(err_val ~ (estimate - uncertainty),
+                                        TRUE ~ (estimate - 1.96 * uncertainty)),
+      estimate_upper = dplyr::case_when(err_val ~ (estimate + uncertainty),
+                                        TRUE ~ (estimate + 1.96 * uncertainty)),
+      fleet = as.character(fleet),
+      year = as.numeric(year)
+    )
+
+  # create plot
+  plt <- ggplot2::ggplot(data = indices2) +
+    ggplot2::geom_line(ggplot2::aes(x = year, y = estimate), linewidth = 1) +
+    ggplot2::geom_ribbon(
+      data = indices2 |> dplyr::filter(!is.na(estimate_lower)),
+      ggplot2::aes(
+        x = year,
+        ymin = estimate_lower,
+        ymax = estimate_upper
+      ),
+      colour = "grey",
+      alpha = 0.3,
+    ) +
+    ggplot2::labs(
+      x = "Year",
+      y = u_units
+    ) +
+    ggplot2::facet_wrap(~ fleet) # +
+    # ggplot2::scale_x_continuous(
+    #   n.breaks = x_n_breaks,
+    #   guide = ggplot2::guide_axis(minor.ticks = TRUE)
+    # )
+
+  plt_fin <- suppressWarnings(add_theme(plt))
+
+  if (make_rda) {
+    # create plot-specific variables to use throughout fxn for naming and IDing
+    topic_label <- "indices_abun"
+
+    # identify output
+    fig_or_table <- "figure"
+
+    # run write_captions.R if its output doesn't exist
+    if (!file.exists(
+      fs::path(getwd(), "captions_alt_text.csv"))
+    ) {
+      satf::write_captions(dat = dat,
+                           dir = rda_dir,
+                           year = end_year)
+    }
+
+    # extract this plot's caption and alt text
+    caps_alttext <- extract_caps_alttext(topic_label = topic_label,
+                                         fig_or_table = fig_or_table)
+
+    export_rda(plt_fin = plt_fin,
+               caps_alttext = caps_alttext,
+               rda_dir = rda_dir,
+               topic_label = topic_label,
+               fig_or_table = fig_or_table)
+  }
+  plt_fin
+}