phytoclass

A pigment-based chemotaxonomic method to determine the biomass of phytoplanton classes.

https://github.com/phytoclass/phytoclass

Repository

A pigment-based chemotaxonomic method to determine the biomass of phytoplanton classes.

README.Rmd

---
output: github_document
---



```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
```

# phytoclass

The phytoclass package uses non-negative matrix factorization and simulated annealing to determine the chl a concentrations of different phytoplankton groups from pigment concentrations.

The main function is 'Simulated annealing'. This can be used in conjuction with the cluster function, which clusters larger datasets. To understand how to use your own matrices instead of the default dataset, please see the example in the vignette.

When setting up your matrices, it is important to ensure that pigments are in the same order in the Sm and Fm matrices. Chlorophyll a must be the final column.

There is a separate function when Divinyl chlorophyll a and Prochlorococcus are used. Here Divinyl chlorophyll a must be the 2nd to last column (see 'Fp').
Note: If dvchl is detected as the second-to-last column in your input matrix, 
the simulated_annealing() function will automatically call the 
Prochlorococcus-optimized version (simulated_annealing_Prochloro()).

The 'Steepest Descent'function is similar to that of CHEMTAX, and is unconstrained by minimum and maximum values, unlike the simualted annealing function. I would use this with more caution.


## Installation

It can be installed from CRAN with:

```r
install.packages("phytoclass")
```

Or the development version from [GitHub](https://github.com/) with:

``` r
# install.packages("devtools")
devtools::install_github("AndyMcKenzieFromNZ/phytoclass")
```

## Example

For the built-in samples matrix Sm of pigment samples:

```{r example}
library(phytoclass)
set.seed("84545")

Results <- simulated_annealing(Sm, niter = 5)

Results$RMSE
Results$MAE
Results$Error
Results$`F matrix`
Results$`Class abudances`
Results$Figure
```


# Documentation
Documentation is built into a website using pkgdown and GitHub actions.
To build locally run `pkgdown::build_site()` and `pkgdown::preview_site()`.
Changes made to the `main` branch will trigger an automatic GitHub action to update the docs.
This process can take a few minutes from commit time to website update.

# Automated Testing
The included `./vignettes/` serve as tests for the project.
If the `pkgdown` build fails then something has gone wrong.

**Always test before pushing** by running `pkgdown::build_site()`.

Owner

GitHub Events

Packages