---
title: "Set up data for multi-species GLMM"
author: "Nathan Brouwer"
date: "February 23, 2018"
output: html_document
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Introduction
This file cleans raw data on MOTU pre/abs and prepares it for regression analysis.

## Preliminaries

### Load libraries

```{r}
library(reshape2)
library(here)
```


### Load raw data

Raw data is in wide format

```{r}
lowa.log.reg <- read.csv("ALL_LOWA_MOTUS_LOG_REG_FAMILY.csv")
```



##  Tidy data 

### Melt data

* Melt data so each MOTU in each sample is on its own line
* Remove empty lines between each row of data
```{r}
glmm.melt <- melt(na.omit(lowa.log.reg),
                         id.vars = names(lowa.nestling.log.reg)[1:7],
                         variable.name = "family")

dim(glmm.melt)
```

### Clean names
```{r}
names(glmm.melt) <- gsub("value","pres.abs",names(glmm.melt))
```



### Center predictor variable

Centering predictors improves GLMM modeling (eg convergence)
```{r}
glmm.melt$PROP_EPT_cent <- scale(glmm.melt$PROPORTION_EPT,
                                                             scale = F)
```

## Save data

```{r}
dat_mult_spp_glmm <- glmm.melt
write.csv(dat_mult_spp_glmm,
          file = "./data_mult_spp_glmm.csv")
```