generate_linear_regression_models.R

require(stringr)
require(pmml)
require(RODBC)


channel <- odbcConnect("PostgreSQL35W32")


#for(houseId in 0:39){
  houseId <- "*"
  
  cat(paste('Create regression model for house_id: ', houseId, '\n'))
  
  results <- sqlQuery(channel, paste("SELECT * FROM smartgird_load_regression_win15min_by_h_fake",houseId))

  pmmlParentXml <- "<PMML version=\"4.2\" xmlns=\"http://www.dmg.org/PMML-4_2\" xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" xsi:schemaLocation=\"http://www.dmg.org/PMML-4_2 http://www.dmg.org/v4-2/pmml-4-2.xsd\">
   <Header copyright=\"Copyright (c) 2014 tom\" description=\"Linear Regression Model\">
    <Extension name=\"user\" value=\"tom\" extender=\"Rattle/PMML\"/>
    <Application name=\"Rattle/PMML\" version=\"1.4\"/>
    <Timestamp>2014-09-01 19:48:16</Timestamp>
   </Header>
   
   
   <DataDictionary numberOfFields=\"2\">
    <DataField name=\"y\" optype=\"continuous\" dataType=\"double\"/>
    <DataField name=\"x\" optype=\"continuous\" dataType=\"double\"/>
   </DataDictionary>
  </PMML>
  "

  parentDoc <<- xmlTreeParse(pmmlParentXml)
  parentDocRootElement <<- xmlRoot(parentDoc)


#sink('d:\\temp\\pmml_output_test_house4_sep8.pmml.xml') 

  computeModel <- function(rec){
    
    house_id <- rec[1]$house_id
    win15minhour <- rec[6]$win15minhour
    
  
    load_dm1 <- rec[9]$load_dm1
    load_dm2 <- rec[10]$load_dm2
    load_dm3 <- rec[11]$load_dm3
    load_dm4 <- rec[12]$load_dm4
    
    if(is.na(load_dm1) | load_dm1 == 0){
      return
    }
  
    days <- c(4,3,2,1)
    loads <- c(load_dm4, load_dm3, load_dm2, load_dm1)
    
    df <- data.frame(x=days, y=loads)
    
    tryCatch({
     model <- lm(formula=y ~ x, data=df)
     #cat(paste('<!-- Model for house_id: ', house_id, ' win15minhour: ', win15minhour, '-->\n',sep=""))
     
     #cat(toString(pmml.lm(model, model.name=paste('load_prediction_h_', house_id,'_win15minhour_',win15minhour, sep=""))))
     pmmlStr <- toString(pmml.lm(model, model.name=paste('load_prediction_h_', house_id,'_win15minhour_',win15minhour, sep="")))
     
     parentDocRootElement <<- addChildren(parentDocRootElement, xmlRoot(xmlTreeParse(pmmlStr))[[3]])
     
     #cat('\n\n')
    }, error = function(e) {
      cat(paste(e,'\n\n'))
    }, finally = {
    })
  }

  by(results, 1:nrow(results), computeModel)
  #sink() 

  saveXML(parentDocRootElement, file=paste("D:\\development\\projects\\xd-keynote\\models\\smartgrid_loadprediction_combined_linregr.pmml.xml", sep=""))
#}

cat("Done!\n")
#parentDocRootElement