README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

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

# quickNmix


The goal of quickNmix is to aid in the fitting of asymptotic N-mixture models, which are computed significantly faster than their canonical counterpart when population sizes are large.

## Installation

You can install the released version of quickNmix from [CRAN](https://CRAN.R-project.org) with:

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

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

``` r
# install.packages("devtools")
devtools::install_github("mrparker909/quickNmix")
```
## Example 1

This is a basic example which shows how to fit a model with site varying lambda, and time varying pdet:

```{r example, cache=T}
library(quickNmix)

tictoc::tic()
nit = anmu[1:2,1:5] # ancient murrelet chick counts
mod = fitNmix(nit=nit,
              K=400, # upper bound on population size
              l_s_c=list(c(0,1)), # lambda site covariate
              p_t_c=list(c(0,1,1,1,1)),
              control=list(reltol=1e-5))
tictoc::toc()
```

model AIC value:

```{r example_p2, cache=T}
mod$model_results$AIC
```

lambda estimates for each site:

```{r example_p3, cache=T}
mod$model_results$estimate_matrices$lambda
```

gamma estimates for each site and time:

```{r example_p4, cache=T}
mod$model_results$estimate_matrices$gamma
```

omega estimates for each site and time:

```{r example_p5, cache=T}
mod$model_results$estimate_matrices$omega
```

pdet estimates for each site and time:

```{r example_p6, cache=T}
mod$model_results$estimate_matrices$pdet
```


## Example 2

This is a basic example which shows how you can use multiple cores to compute:

```{r example2, cache=T}
library(quickNmix)

# uses library doParallel, here we use 2 cores
doParallel::registerDoParallel(cores = 2) 

tictoc::tic()
nit  = anmu[c(2,5),1:12] # ancient murrelet chick counts
mod2 = fitNmix(nit=nit, K=400)
tictoc::toc()

mod2
```


Multi-threading is implemented using R packages doParallel and foreach. Note that multi-threading is used to split the computation of the transition probability matrix by rows. This may not be efficient on some architectures, and will not be efficient for small K: efficiency increases with increasing K. Alternative choices for multi-core processing are being considered.