I have an excel file of rules: a column for Antecedents, another for Consequents, other for Support, etc
I'm trying to create a "rules" object to plot them using arulesViz.
http://www.inside-r.org/packages/cran/arules/docs/rhs
new("rules", ...)
I'm trying to create the Slots
lhs:Object of class itemMatrix; the left-hand-sides of the rules (antecedents)
rhs:Object of class itemMatrix; the right-hand-sides of the rules (consequents)
quality:a data.frame
To create the itemMatrix I need to go back to the sparse matrix,but I think that won't work, Is there some way to "import" the rules to arulesViz?
Excel file:
Antec Conseq Supp Conf
MMMMAAA MMAAAA 0.061945 0.5
MMM,MA MMAAAA 0.071944 0.6
MMMMAAA MMAAA 0.053948 0.5
MMM,MA MMAAA 0.054948 0.7
AAAAAA AAAA 0.090909 0.5
One way would be to create a PMML file for the rules and use read.PMML. To create a rules object from scratch is a little tricky. Here is an example:
library("arules")
l <- list(c("MMMMAAA"), c("MMM", "MA"), "MMMMAAA", c("MMM","MA"), "AAAAA")
r <- list("MMAAAA", "MMAAAA", "MMAAA", "MMAAA", "AAAA")
q <- data.frame(
support = c(0.061945, 0.071944, 0.053948, 0.054948,0.090909),
confidence = c(.5, .6, .5, .7, .5),
lift = c(1,1,1,1,1)
)
### Note that I also added lift since arulesViz uses
### lift in some visualizations.
### find unique item labels
items <- unique(c(unlist(l), unlist(r)))
### encode data as a rules object
r <- new("rules", lhs = encode(l, items),
rhs = encode(r, items), quality = q)
inspect(r)
### use a visualization as a scatter plot.
library("arulesViz")
plot(r)
Hope this helps!