In laurieKell/FLCandy: Candidate methods for FLR

knitr::opts_chunk$set(echo = TRUE, warning = FALSE, message = FALSE)

Introduction

This vignette demonstrates how to analyze weights-at-age data using FLR (Fisheries Library in R) methods. We'll cover summarizing, visualizing, and simulating weights-at-age data using FLQuant objects.

Required Packages

library(FLCore)
library(ggplot2)
library(plyr)
library(dplyr)
library(corrplot)

source("C:/active/FLCandy/examples/var/ts.R")

Loading and Preparing Data

First, let's load our weights-at-age data from the CSV file and convert it to an FLQuant object:

data(ple4)

waa=stock.wt(ple4)

Summarizing Weights-at-Age

We'll use the tsSmry function to calculate summary statistics:

summary=tsSmry(waa)
summary(summary)

Visualizing Weights-at-Age

Now let's create some visualizations of our data:

# Time series plot
ggplot(as.data.frame(waa), aes(x = year, y = data, color = factor(age))) +
  geom_line() +
  labs(title = "Weights-at-Age Over Time", x = "Year", y = "Weight", color = "Age") +
  theme_minimal()

# Boxplot of weights by age
ggplot(as.data.frame(waa), aes(x = factor(age), y = data)) +
  geom_boxplot() +
  labs(title = "Distribution of Weights by Age", x = "Age", y = "Weight") +
  theme_minimal()

# Heatmap of weights-at-age over time
ggplot(as.data.frame(waa), aes(x = year, y = factor(age), fill = data)) +
  geom_tile() +
  scale_fill_viridis_c() +
  labs(title = "Heatmap of Weights-at-Age Over Time", x = "Year", y = "Age", fill = "Weight") +
  theme_minimal()

Analyzing Correlations

Let's examine the correlations between age groups:

cor_matrix <- tsCor(waa)[,,drop=TRUE]
corrplot(cor_matrix, method = "color", type = "upper", order = "hclust",
         tl.col = "black", tl.srt = 45, addCoef.col = "black")

Autocorrelation Analysis

We can analyze the autocorrelation in the time series:

acf_results <- tsACF(waa)
ggplot(as.data.frame(acf_results), aes(x = year, y = data, color = factor(age))) +
  geom_line() +
  labs(title = "Autocorrelation Function by Age", x = "Lag", y = "ACF", color = "Age") +
  theme_minimal()

Generating Correlated Random Deviates

Finally, let's generate correlated random deviates based on the original data:

n_simulations <- 100
simulated_deviates <- deviates(n_simulations, waa, tsSmry(waa)[["var"]], tsCov(waa)[,,drop=TRUE])

# Visualize the first simulation
ggplot(as.data.frame(simulated_deviates[,,,,,1]), aes(x = year, y = data, color = factor(age))) +
  geom_line() +
  labs(title = "Simulated Weights-at-Age (First Simulation)", x = "Year", y = "Weight", color = "Age") +
  theme_minimal()

Conclusion

This vignette demonstrated how to analyze weights-at-age data using FLR methods. We covered data loading, summarization, visualization, correlation analysis, and simulation of new data based on the original data's properties. These techniques can be valuable for understanding the dynamics of fish populations and for use in stock assessment models.

laurieKell/FLCandy documentation built on April 17, 2025, 5:23 p.m.