Code for demos from H2O in R meetup. Must have latest R package (git …

…pull, make) to run the data munging part!

R/examples/H2OMeetupDemo.R

@@ -0,0 +1,29 @@
+library(h2o)
+localH2O = new("H2OClient")
+h2o.checkClient(localH2O)
+
+# For hands-on audience participation
+# H2O Import, Summary, GLM and K-Means on prostate cancer data set
+# prostate.hex = h2o.importFile(localH2O, path = "../../smalldata/logreg/prostate.hex", key = "prostate.hex")
+prostate.hex = h2o.importURL(localH2O, path = "https://raw.github.com/0xdata/h2o/master/smalldata/logreg/prostate.csv", key = "prostate.hex")
+summary(prostate.hex)
+prostate.glm = h2o.glm(y = "CAPSULE", x = c("AGE","RACE","PSA","GLEASON"), data = prostate.hex, family = "binomial", nfolds = 10, alpha = 0.5)
+print(prostate.glm)
+prostate.km = h2o.kmeans(data = prostate.hex, centers = 5, cols = c("AGE","RACE","GLEASON","CAPSULE","PSA"))
+print(prostate.km)
+
+# Still in Beta! H2O Data Munging on prostate cancer data set
+head(prostate.hex, n = 10)
+tail(prostate.hex)
+summary(prostate.hex$AGE)
+summary(prostate.hex[prostate.hex$AGE > 67,])
+prostate.small = as.data.frame(prostate.hex[1:200,])
+glm(CAPSULE ~ AGE + RACE + DPROS + DCAPS, family = binomial, data = prostate.small)
+
+# R Import, Summary, GLM and K-Means on prostate cancer data set
+prostate.data = read.csv(url("https://raw.github.com/0xdata/h2o/master/smalldata/logreg/prostate.csv"), header = TRUE)
+summary(prostate.data)
+prostate.glm2 = glm(CAPSULE ~ AGE + RACE + PSA + GLEASON, family = binomial, data = prostate.data)
+print(prostate.glm2)
+prostate.km2 = kmeans(prostate.data[c("AGE","RACE","GLEASON","CAPSULE","PSA")], centers = 5)
+print(prostate.km2)

R/examples/H2OMeetupDemo2.R

@@ -0,0 +1,29 @@
+library(h2o)
+localH2O = new("H2OClient")
+h2o.checkClient(localH2O)
+
+# For hands-off demo of H2O vs. R
+# H2O Import, Summary and GLM of small airlines data set on local machine
+airlines.hex = h2o.importURL(localH2O, path = "https://raw.github.com/0xdata/h2o/master/smalldata/airlines/allyears2k_headers.zip", key = "airlines.hex")
+summary(airlines.hex)
+x_ignore = c("IsArrDelayed", "ActualElapsedTime", "ArrDelay", "DepDelay", "Canceled", "Diverted", "IsDepDelayed")
+myX = setdiff(colnames(airlines.hex), x_ignore)
+airlines.glm = h2o.glm(y = "IsArrDelayed", x = myX, data = airlines.hex, family = "binomial", nfolds = 10, alpha = 0.5)
+print(airlines.glm)
+
+# For hands-on demo of running H2O remotely
+# H2O Import, Summary and GLM of large airlines data set on remote machine
+remoteH2O = new("H2OClient", ip="192.168.1.161", port=54329)
+h2o.checkClient(remoteH2O)
+
+airlines_big.hex = h2o.importFile(remoteH2O, path = "/home/earl/./oldairlines/airlines.orig.all.withheader.25.csv", key = "airlines_big.hex")
+summary(airlines_big.hex)
+airlines_big.glm = h2o.glm(y = "IsArrDelayed", x = myX, data = airlines_big.hex, family = "binomial", nfolds = 10, alpha = 0.5)
+print(airlines_big.glm)
+
+# Still in Beta! H2O Data Munging on large airlines data set
+head(airlines_big.hex, n = 10)
+tail(airlines_big.hex)
+summary(airlines_big.hex$ArrDelay)
+airlines_small.data = as.data.frame(airlines_big.hex[1:10000,])
+glm(IsArrDelayed ~ Origin + Dest, family = binomial, data = airlines_small.data)