FP-growth
spark.fpGrowth.RdA parallel FP-growth algorithm to mine frequent itemsets.
spark.fpGrowth fits a FP-growth model on a SparkDataFrame. Users can
spark.freqItemsets to get frequent itemsets, spark.associationRules to get
association rules, predict to make predictions on new data based on generated association
rules, and write.ml/read.ml to save/load fitted models.
For more details, see
FP-growth.
Usage
spark.fpGrowth(data, ...)
spark.freqItemsets(object)
spark.associationRules(object)
# S4 method for class 'SparkDataFrame'
spark.fpGrowth(
data,
minSupport = 0.3,
minConfidence = 0.8,
itemsCol = "items",
numPartitions = NULL
)
# S4 method for class 'FPGrowthModel'
spark.freqItemsets(object)
# S4 method for class 'FPGrowthModel'
spark.associationRules(object)
# S4 method for class 'FPGrowthModel'
predict(object, newData)
# S4 method for class 'FPGrowthModel,character'
write.ml(object, path, overwrite = FALSE)Arguments
- data
A SparkDataFrame for training.
- ...
additional argument(s) passed to the method.
- object
a fitted FPGrowth model.
- minSupport
Minimal support level.
- minConfidence
Minimal confidence level.
- itemsCol
Features column name.
- numPartitions
Number of partitions used for fitting.
- newData
a SparkDataFrame for testing.
- path
the directory where the model is saved.
- overwrite
logical value indicating whether to overwrite if the output path already exists. Default is FALSE which means throw exception if the output path exists.
Value
spark.fpGrowth returns a fitted FPGrowth model.
A SparkDataFrame with frequent itemsets.
The SparkDataFrame contains two columns:
items (an array of the same type as the input column)
and freq (frequency of the itemset).
A SparkDataFrame with association rules.
The SparkDataFrame contains five columns:
antecedent (an array of the same type as the input column),
consequent (an array of the same type as the input column),
confidence (confidence for the rule)
lift (lift for the rule)
and support (support for the rule)
predict returns a SparkDataFrame containing predicted values.
Note
spark.fpGrowth since 2.2.0
spark.freqItemsets(FPGrowthModel) since 2.2.0
spark.associationRules(FPGrowthModel) since 2.2.0
predict(FPGrowthModel) since 2.2.0
write.ml(FPGrowthModel, character) since 2.2.0
Examples
if (FALSE) { # \dontrun{
raw_data <- read.df(
"data/mllib/sample_fpgrowth.txt",
source = "csv",
schema = structType(structField("raw_items", "string")))
data <- selectExpr(raw_data, "split(raw_items, ' ') as items")
model <- spark.fpGrowth(data)
# Show frequent itemsets
frequent_itemsets <- spark.freqItemsets(model)
showDF(frequent_itemsets)
# Show association rules
association_rules <- spark.associationRules(model)
showDF(association_rules)
# Predict on new data
new_itemsets <- data.frame(items = c("t", "t,s"))
new_data <- selectExpr(createDataFrame(new_itemsets), "split(items, ',') as items")
predict(model, new_data)
# Save and load model
path <- "/path/to/model"
write.ml(model, path)
read.ml(path)
# Optional arguments
baskets_data <- selectExpr(createDataFrame(itemsets), "split(items, ',') as baskets")
another_model <- spark.fpGrowth(data, minSupport = 0.1, minConfidence = 0.5,
itemsCol = "baskets", numPartitions = 10)
} # }