---
title: "Using the MUMARINEX Package"
author: "Nathan Chauvel"
ate: "`r Sys.Date()`"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Using the MUMARINEX Package}
  %\VignetteEncoding{UTF-8}
  %\VignetteEngine{knitr::rmarkdown}
editor_options: 
  chunk_output_type: console
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

# Introduction

The ***mumarinex*** package provides tools for the computation of the **MUltivariate MArine Recovery INdEX (MUMARINEX)** as described in Chauvel *et al.* (2026). This index is designed to evaluate community recovery in marine ecosystems by combining three complementary sub-indices:

-   **Standardized Corrected Shared Richness (SCSR)**: community richness changes (taxa loss or gain)
-   **Corrected Bray-Curtis Similarity (CBCS)**: community structure changes (decrease or increase in abundance)
-   **Standardized Pielou Index (SPI)**: community dominance shifts (appearance or changes in dominance)

The package also includes diagnostic and visualization functions to identify which taxa or ecological mechanisms drive the observed variations.

In this vignette, we will observe how to:

1.  Install and load the package.
2.  Format data.
3.  Compute the MUMARINEX index and its sub-indices (SCSR, CBCS and SPI).
4.  Visualize sub-indices results with dedicated plotting functions.
5.  Generate diagnostics to interpret sub-indices variations.

All examples below use the dataset **`Simulated_data`**, included in the package.\
For details on how this dataset was constructed, please refer to its documentation page (`?Simulated_data`)

# Install and load the package

You can install the development version of mumarinex from CRAN or from [GitHub](https://github.com/), respectively with:

```r load-data, message=FALSE

install.packages("mumarinex") # CRAN
devtools::install_github("Nathan-Chauvel/mumarinex") # GitHub
```
Once you run one of this above command, you can load the package with:
```{r}
library(mumarinex)
```


# Data format

The input data must be provided as a data frame or matrix, with rows representing samples and columns representing species. A reference vector specifying the reference samples must also be supplied. The formatting of Simulated_data can be used as a template for preparing your own dataset. In this example, the reference stations (REF1, REF2) are located in rows 1 to 10. The example dataset can be loaded into the R environment as follows:

```{r}
# Load example dataset
data("Simulated_data")

# Display the first rows
head(Simulated_data)

# Definition of the reference position
ref_idx <- 1:10 # row number of the reference samples
```

# Compute the MUMARINEX index and its sub-indices (SCSR, CBCS and SPI)

Once the data are properly defined, the MUMARINEX index and its sub-indices can be computed. The function `mumarinex()` calculates the MUMARINEX index, and by setting `subindices = TRUE`, it also returns the three complementary sub-indices.

```{r}
# Compute MUMARINEX and sub-indices
rMUM <- mumarinex(x = Simulated_data, ref = ref_idx, subindices = TRUE, log = FALSE)

# Extract MUMARINEX
rMUMARINEX<-rMUM$MUMARINEX

# Extract sub-indices
Subind<-rMUM$Subindices
```

MUMARINEX results can subsequently be examined through graphical representations, such as boxplots.

```{r,fig.width=7, fig.height=5}
stations<-matrix(unlist(strsplit(rownames(Simulated_data),".",fixed=TRUE)),ncol=2,byrow=TRUE)[,1] # get station labels from data rownames

stations<-factor(stations,levels=unique(stations)) # setting station names as factor to specify in which order it must display it in the boxplot

boxplot(rMUMARINEX~stations,ylim=c(0,1)) # ylim is set in the interval 0-1 as it is the range of MUMARINEX
```

# Visualize sub-indices results with dedicated plotting functions

To better understand the variations in MUMARINEX, it is often useful to examine how the sub-indices vary. The `decomplot()` function displays the distribution of these sub-indices (SCSR, CBCS, SPI) across sample groups using boxplots.

```{r,fig.width=10, fig.height=5}

decomplot(x = Simulated_data, g = stations, ref = ref_idx, log = FALSE, main = "Artificial data")
```

# Generate diagnostics to interpret sub-indices variations

Once the sub-index variations underlying the final MUMARINEX value have been examined, the `diagnostic_tool()` function can be used to identify the species that best account for these changes.

```{r}
diagnostic_tool(x = Simulated_data, g = stations, ref = ref_idx, log = FALSE)
```