Autocorrelation testing

##Autocorrelation testing with Starry Stonewort Data 2021

library(ggplot2)
library(dplyr)
library(tidyr)
library(pspearman)
library(sp)
library(tseries)

# We have two different sets of data that might be autocorrelated so we're gonna test both

SSWnomod <- read.csv("SSW_Data.csv")
SSWphmod <- read.csv("SSW_Data_Final_pH_Modified.csv")

# First let's get the data sheet cut up into the variables we want
## We're going to remove the organizational measure (first 5 columns)
# Data needs to have all values if a value is missing then put an NA

newnomod <- SSWnomod[,6:14]
newmod <- SSWphmod[,6:14]

Autonomod <- acf(newnomod, type= c("correlation"))
Automod <- acf(newmod, type = c("correlation"))

## Tons of plots are going to pop up so go through them piece by piece and save them as images if you want
## we essentially want to be sure that our data has a degree of randomness to it
## Lag at 0 will always be 1 since that will be the data tested against itself 
# the dotted lines will be a 95% confidence band on the data so if your lag at the other values is above that 
#then we would see that the data has some autocorrelation present

# For the SSW data we can see that the tissues are autocorrelated with one another for the most part but
# between the tissues and the environmental variables there isn't too much autocorrelation 
# and between the environmental variables they are decently autocorrelated as well