Temp Scripts/xVignette1_SampleData_TestPieces.R

In SOLV-Code/SPATFunctions-Package: Functions for the Salmon Pattern Analysis Tool

# BUILDING BLOCKS FOR VIGNETTE 1: SAMPLE-LEVEL DATA

# load the package
library(SPATFunctions)


# prep a subset of the data

test.data <- SPATData_TimeSeries %>% mutate(RpS = Recruits/Spawners) %>%
  select(Stock,Year,RpS) %>%
  pivot_wider(names_from = Stock,values_from = RpS) %>% select(-Year)

test.years <- sort(unique(SPATData_TimeSeries$Year))

head(test.data)
dim(test.data)
length(test.years)









# sources
# https://anomaly.io/detect-correlation-time-series/index.html
# https://www.r-bloggers.com/tidy-time-series-analysis-part-3-the-rolling-correlation/




library(SPATFunctions)
library(tidyverse)
library(tidyr)


# prep some data


test.data <- SPATData_TimeSeries %>% mutate(RpS = Recruits/Spawners) %>%
  select(Stock,Year,RpS) %>%
  pivot_wider(names_from = Stock,values_from = RpS) %>% select(-Year)

test.years <- sort(unique(SPATData_TimeSeries$Year))

head(test.data)
dim(test.data)
length(test.years)


####################################################################
# BASE R FUNCTIONS
####################################################################


# basic correlation and covariation matrices
# all time series at once
# but does not work with NA
# -> need to drop all years where any 1 obs is missing!
# -> major problem

cov(na.omit(test.data))
cor(na.omit(test.data))


# do pairwise cross-correlation or cross-variance
# -> could set this up to work through all pairs and produce a matrix
# -> how would this differ from basic Cov(), Cor()? -> would use different subsets
# depending on complete pairs, rather tham years that are complete for ALL stocks

tmp.out <- ccf(test.data$Stock1, test.data$Stock2, lag.max = 5, type = "correlation",
               plot = TRUE, na.action = na.omit)
tmp.out



###########################################################################
# TTR Package
#########################################################################
install.packages("TTR")
library(TTR)

data.use <- na.omit(select(test.data,Stock1, Stock2))
na.idx <-  !complete.cases(select(test.data,Stock1, Stock2))
test.fit.cov <- runCov(data.use[,1],data.use[,2], n = 10, sample = TRUE, cumulative = TRUE)
test.fit.cor <- runCor(data.use[,1],data.use[,2], n = 10, sample = TRUE, cumulative = TRUE)

plot(test.years[!na.idx], test.fit.cov,type="l")
plot(test.years[!na.idx], test.fit.cor,type="l")












######################################################
# corrr package
#####################################################

install.packages("corrr")
library(corrr)

test.out <- correlate(iris[-5])
network_plot(test.out)

# not working
#corrplot(test.out,is.corr=FALSE)



##########################################
# corrplot package
###################################


library(corrplot)

M <- cor(mtcars)
corrplot(M)

corrplot(M, method = "number",
         col = "black", cl.pos = "n")
corrplot(M, method = "number")
corrplot(M, order = "AOE")
corrplot(M, method="color",order = "hclust")
corrplot(M, order = "AOE",
         addCoef.col = "grey")


##########################################
#roll_cor: Rolling Correlation Matrices
########################################









### Other Packages

# xts, zoo,