#!/usr/bin/env Rscript

if (!require(arules)) install.packages("arules", repos = "https://cran.r-project.org/", Ncpus = 16) # nolint
library(arules)
if (!require(arulesViz)) install.packages("arulesViz", repos = "https://cran.r-project.org/", Ncpus = 16) # nolint
library(arulesViz)

t <- read.transactions("AssociationRules.csv", format = "basket", sep = " ")

summary(t)

# a most freq
names(sort(itemFrequency(t), decreasing = TRUE)[1])
# b max amount of items
max(size(t))


# 0..0.01 support, 0 confidence
rules0 <- apriori(t, parameter = list(supp = 0.01, conf = 0, minlen = 2))
# c
length(rules0)

# d
rules05 <- apriori(t, parameter = list(supp = 0.01, conf = 0.5, minlen = 2))
length(rules05)

# f
plot(rules05, measure = c("support", "lift"), shading = "confidence")
# g
plot(rules05, measure = c("support", "confidence"), shading = "lift")

# j
rules_supp_01 <- subset(rules05, support >= 0.1)
rules_by_support <- sort(rules_supp_01, by = "support", descreasing = FALSE)
inspect(rules_by_support)


# k
rules_conf_08 <- sort(subset(rules05, confidence > 0.8), by = "lift", descreasing = FALSE)
inspect(rules_conf_08)

plot(rules_conf_08,
  method = "matrix",
  engine = "grid",
  measure = "confidence",
  shading = "lift",
  control = list(recorded = FALSE)
)

plot(rules_conf_08,
  method = "matrix",
  engine = "grid",
  shading = c("lift", "confidence")
)


# n
rules_lift_max_3 <- head(sort(rules05, by = "lift", descreasing = TRUE), 3)
# o
plot(rules_lift_max_3, method = "graph", engine = "igraph")
inspect(rules_lift_max_3)

exit
# q
train_t <- head(t, 8000)
test_t <- tail(t, 2000)

train <- apriori(train_t, parameter = list(supp = 0.01, conf = 0.5))
subset_train <- subset(train, lift > 15)
subset_train_df <- as(subset_train, "data.frame")

# https://www.rdocumentation.org/packages/arules/versions/1.7-9/topics/interestMeasure
test <- interestMeasure(subset_train, transactions = test_t, measure = c("support", "confidence", "lift", "count"), reuse = FALSE)

for (i in 1:length(subset_train)) {
  cat(subset_train_df$rules[i])
  cat("\n")
  cat("train conf:", subset_train_df$confidence[i], "lift:", subset_train_df$lift[i])
  cat("\n")
  cat("test  conf:", test$confidence[i], "lift:", test$lift[i])
  cat("\n")
}