[SPARK-13449] Naive Bayes wrapper in SparkR

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.0.0
• Component/s: ML, SparkR
• Target Version/s: 2.0.0
• Assignee: Xusen Yin
• Reporter: Xiangrui Meng
• Shepherd: Yanbo Liang
• Created: 23/Feb/16 08:19
• Updated: 22/Mar/16 21:17
• Resolved: 22/Mar/16 21:17

Description

Following SPARK-13011, we can add a wrapper for naive Bayes in SparkR. R's naive Bayes implementation is from package e1071 with signature:

## S3 method for class 'formula'
naiveBayes(formula, data, laplace = 0, ..., subset, na.action = na.pass)
## Default S3 method, which we don't want to support
# naiveBayes(x, y, laplace = 0, ...)


## S3 method for class 'naiveBayes'
predict(object, newdata,
  type = c("class", "raw"), threshold = 0.001, eps = 0, ...)

It should be easy for us to match the parameters.

Issue Links

• relates to: SPARK-13641 getModelFeatures of ml.api.r.SparkRWrapper cannot (always) reveal the original column names
• relates to: SPARK-13011 K-means wrapper in SparkR
• links to: [Github] Pull Request #11486 (yinxusen)
• links to: [Github] Pull Request #11890 (mengxr)