---
title: "Example code"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{demo-code}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

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

## Instructions to run a demo

To download a set of demo files, run the following code.
```
fpath <- system.file("extdata", package="NWCTrends")
file.copy(fpath, ".",recursive=TRUE)
```
This will create a folder called `extdata` in your directory.

To run a demo analysis and create a report, type 
```
library(NWCTrends)
NWCTrends_report()
```
You will be asked to select a data file. If you do not have data, navigate to one of the .csv or .RData files in the `extdata` folder.

Type `?NWCTrends` for instructions for analyzing a data set. The data must be .csv file. Figures will be saved in the **NWCTrends_output** folder, created in your working directory.  

## Instructions to run your own data

Download the demo data files and duplicate the format (csv or xls). Do not rename the columns. Missing data are entered with a -99. An ESU name and a unique population name (COMMON_POPULATION_NAME) are required. The tables produced by **NWCTrends** only use the BROOD_YEAR, NUMBER_OF_SPAWNERS, and FRACWILD columns. An entry is required for each year. RUN_TIME, SPECIES and MAJOR_POPULATION_GROUP are used to adding labels to plots and tables. BROOD_YEAR is mislabelled. This is simply the Year of sampling.

## Modifying the tables

Optional. Set the years to use for the analysis using `fit.min.year` and `fit.max.year`. If you leave this off, it will use the first year in the data set and the last year in the data set.

In the tables for the geometric means, you can control the table by passing in the list `geomean.table.control`. For example, you can set the beginning and ending years to be shown in the table. These can be different than `fit.min.year` and `fit.max.year`. 

For example, you may want to fit to 1990 to 2019 data but only show
5-year geometric means for 1999 to 2018. To do this, you would call the report with

```
NWCTrends_report(fit.min.year=1990, fit.max.year=2019,
                 geomean.table.control=list(min.year=1999, max.year=2018, change.col="first.last"))
```
The `change.col` argument determines whether the last column is the percent change between the first and last 5-year bands or between the last two bands.

The code will create bands with 5 years in each band starting with min.year. If max.year,
would lead to a final band with less than 5 years, then the last band will not have 5
years. If it has fewer than min.band.points, then the last band will be NA.
You will need to properly choose min.year and max.year to get the table to look as you want.

You customize the multi-year trend ranges. To do this, you would call the report with for example the following to show the 1990 to 2019 and 2015 to 2019 ranges.

```
NWCTrends_report(fit.min.year=1990, fit.max.year=2019,
                 trend.table.control=list(year.ranges=list(1990:2019, 2015:2019)))
```