data/scripts/augment_meteo.R

   1 meteo_df = read.csv("meteo.csv")
   2
   3 #inspiration
   4 #http://stackoverflow.com/questions/8214303/conditional-replacement-of-values-in-a-data-frame
   5
   6 meteo_df$Season = 0
   7 meteo_df$Week = 0
   8 meteo_df$Pollution = -1
   9
  10 #Need to load and aggregate PM10 by days: use getData() from package
  11 data = getData(..., predict_at=0) #TODO:
  12
  13 for (i in 1:nrow(meteo_df))
  14 {
  15     pm10_level = data$getLevel(i)
  16     #Fill Pollution column: -1 if no info, 0 to 2 for pollution level
  17     if (!is.nan(pm10_level))
  18     {
  19         if (pm10_level < 30)
  20             meteo_df$Pollution[i] = 0
  21         else if (pm10_level <= 50)
  22             meteo_df$Pollution[i] = 1
  23         else #pm10 > 50
  24             meteo_df$Pollution[i] = 2
  25     }
  26
  27     #Also fill season + days of week variables
  28     meteo_df$Season[i] = ifelse(
  29         strsplit(as.character(meteo_df$Date[i]),'/')[[1]][1] %in% c("4","5","6","7","8"),
  30         1, 0)
  31     current_datetime = strptime(as.character(meteo_df$Date[i]), "%m/%d/%Y", tz="GMT")
  32     meteo_df$Week[i] = ifelse(current_datetime$wday %in% c(6,0), 0, 1)
  33 }
  34
  35 #Finally write new data
  36 write.csv(meteo_df, file="meteo_extra.csv", row.names=FALSE)