Skip to content

Commit

Permalink
Updated test resources
Browse files Browse the repository at this point in the history
  • Loading branch information
@vruusmann
vruusmann committed May 14, 2019
1 parent a71d9e7 commit 0b8dade
Show file tree
Hide file tree
Showing 13 changed files with 12,842 additions and 11,692 deletions.
38 changes: 19 additions & 19 deletions src/test/R/xgboost.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ insertNA = function(df){
# Regression
#

predictAutoMpg = function(auto.xgb, auto.dmatrix, ntreelimit = NULL){
predictAutoMpg = function(auto.xgb, auto.dmatrix, ntreelimit){
mpg = predict(auto.xgb, newdata = auto.dmatrix, ntreelimit = ntreelimit)

return (data.frame("_target" = mpg))
Expand All @@ -62,7 +62,7 @@ genAutoMpg = function(auto_y, auto_X, dataset, ...){
auto.xgb = xgboost(data = auto.dmatrix, objective = "reg:linear", nrounds = 31, ...)

storeModel(auto.xgb, funcAndDataset, dataset)
storeResult(predictAutoMpg(auto.xgb, auto.dmatrix), funcAndDataset)
storeResult(predictAutoMpg(auto.xgb, auto.dmatrix, 31), funcAndDataset)
}

auto = loadCsv("csv/Auto.csv")
Expand All @@ -86,7 +86,7 @@ genAutoMpg(auto_y, auto_X, "AutoNA")
# Poisson, Tweedie and Gamma regressions
#

predictVisitCount = function(visit.xgb, visit.dmatrix, ntreelimit = NULL){
predictVisitCount = function(visit.xgb, visit.dmatrix, ntreelimit){
count = predict(visit.xgb, newdata = visit.dmatrix, ntreelimit = ntreelimit)

return (data.frame("_target" = count))
Expand All @@ -105,7 +105,7 @@ genVisitCount = function(visit_y, visit_X, dataset, ...){
visit.xgb = xgboost(data = visit.dmatrix, objective = "count:poisson", nrounds = 31, ...)

storeModel(visit.xgb, funcAndDataset, dataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix), funcAndDataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix, 31), funcAndDataset)

funcAndDataset = paste("TweedieRegression", dataset, sep = "")

Expand All @@ -114,7 +114,7 @@ genVisitCount = function(visit_y, visit_X, dataset, ...){
visit.xgb = xgb.train(data = visit.dmatrix, params = list(objective = "reg:tweedie", tweedie_variance_power = 1.5), nrounds = 31)

storeModel(visit.xgb, funcAndDataset, dataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix), funcAndDataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix, 31), funcAndDataset)

funcAndDataset = paste("GammaRegression", dataset, sep = "")

Expand All @@ -123,7 +123,7 @@ genVisitCount = function(visit_y, visit_X, dataset, ...){
visit.xgb = xgboost(data = visit.dmatrix, objective = "reg:gamma", nrounds = 31, ...)

storeModel(visit.xgb, funcAndDataset, dataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix), funcAndDataset)
storeResult(predictVisitCount(visit.xgb, visit.dmatrix, 31), funcAndDataset)
}

visit = loadCsv("csv/Visit.csv")
Expand All @@ -145,13 +145,13 @@ genVisitCount(visit_y, visit_X, "VisitNA")
# Binary classification
#

predictBinomialAuditAdjusted = function(audit.xgb, audit.dmatrix, ntreelimit = NULL){
predictBinomialAuditAdjusted = function(audit.xgb, audit.dmatrix, ntreelimit){
probability = predict(audit.xgb, newdata = audit.dmatrix, ntreelimit = ntreelimit)

return (data.frame("_target" = as.integer(probability > 0.5), "probability(0)" = (1 - probability), "probability(1)" = probability, check.names = FALSE))
}

predictMultinomialAuditAdjusted = function(audit.xgb, audit.dmatrix, ntreelimit = NULL){
predictMultinomialAuditAdjusted = function(audit.xgb, audit.dmatrix, ntreelimit){
probability = predict(audit.xgb, newdata = audit.dmatrix, ntreelimit = ntreelimit, reshape = TRUE)

return (data.frame("_target" = as.integer(probability[, 2] > 0.5), "probability(0)" = probability[, 1], "probability(1)" = probability[, 2], check.names = FALSE))
Expand All @@ -169,7 +169,7 @@ genAuditAdjusted = function(audit_y, audit_X, dataset, ...){

audit.xgb = xgboost(data = audit.dmatrix, objective = "reg:logistic", nrounds = 17, ...)

adjusted = predict(audit.xgb, newdata = audit.dmatrix)
adjusted = predict(audit.xgb, newdata = audit.dmatrix, ntreelimit = 17)

storeModel(audit.xgb, funcAndDataset, dataset)
storeResult(data.frame("_target" = adjusted), funcAndDataset)
Expand All @@ -178,19 +178,19 @@ genAuditAdjusted = function(audit_y, audit_X, dataset, ...){

set.seed(42)

audit.xgb = xgboost(data = audit.dmatrix, objective = "binary:logistic", nrounds = 71)
audit.xgb = xgboost(data = audit.dmatrix, objective = "binary:logistic", nrounds = 71, ...)

storeModel(audit.xgb, funcAndDataset, dataset)
storeResult(predictBinomialAuditAdjusted(audit.xgb, audit.dmatrix), funcAndDataset)
storeResult(predictBinomialAuditAdjusted(audit.xgb, audit.dmatrix, 71), funcAndDataset)
storeResult(predictBinomialAuditAdjusted(audit.xgb, audit.dmatrix, 31), paste(funcAndDataset, "31", sep = "@"))

funcAndDataset = paste("MultinomialClassification", dataset, sep = "")

set.seed(42)

audit.xgb = xgboost(data = audit.dmatrix, objective = "multi:softprob", num_class = 2, nrounds = 31)
audit.xgb = xgboost(data = audit.dmatrix, objective = "multi:softprob", num_class = 2, nrounds = 31, ...)
storeModel(audit.xgb, funcAndDataset, dataset)
storeResult(predictMultinomialAuditAdjusted(audit.xgb, audit.dmatrix), funcAndDataset)
storeResult(predictMultinomialAuditAdjusted(audit.xgb, audit.dmatrix, 31), funcAndDataset)
}

audit = loadCsv("csv/Audit.csv")
Expand All @@ -213,7 +213,7 @@ genAuditAdjusted(audit_y, audit_X, "AuditNA")
# Multi-class classification
#

predictIrisSpecies = function(iris.xgb, iris.dmatrix, ntreelimit = NULL){
predictIrisSpecies = function(iris.xgb, iris.dmatrix, ntreelimit){
probabilities = predict(iris.xgb, newdata = iris.dmatrix, ntreelimit = ntreelimit)

# Convert from vector to three-column matrix
Expand All @@ -224,7 +224,7 @@ predictIrisSpecies = function(iris.xgb, iris.dmatrix, ntreelimit = NULL){
return (data.frame("_target" = species, "probability(0)" = probabilities[, 1], "probability(1)" = probabilities[, 2], "probability(2)" = probabilities[, 3], check.names = FALSE))
}

genIrisSpecies = function(iris_y, iris_X, dataset){
genIrisSpecies = function(iris_y, iris_X, dataset, ...){
iris.fmap = genFMap(iris_X)
writeFMap(iris.fmap, csvFile(dataset, ".fmap"))

Expand All @@ -234,11 +234,11 @@ genIrisSpecies = function(iris_y, iris_X, dataset){

set.seed(42)

iris.xgb = xgboost(data = iris.dmatrix, objective = "multi:softprob", num_class = 3, nrounds = 17)
iris.xgb = xgboost(data = iris.dmatrix, objective = "multi:softprob", num_class = 3, nrounds = 17, ...)

storeModel(iris.xgb, funcAndDataset, dataset)
storeResult(predictIrisSpecies(iris.xgb, iris.dmatrix), funcAndDataset)
storeResult(predictIrisSpecies(iris.xgb, iris.dmatrix, ntreelimit = 11), paste(funcAndDataset, "11", sep = "@"))
storeResult(predictIrisSpecies(iris.xgb, iris.dmatrix, 17), funcAndDataset)
storeResult(predictIrisSpecies(iris.xgb, iris.dmatrix, 11), paste(funcAndDataset, "11", sep = "@"))
}

iris = loadCsv("csv/Iris.csv")
Expand All @@ -249,7 +249,7 @@ iris_X = iris[, 1:(ncol(iris) - 1)]
# Convert from factor levels to 0-based indexes
iris_y = (as.integer(iris_y) - 1)

genIrisSpecies(iris_y, iris_X, "Iris")
genIrisSpecies(iris_y, iris_X, "Iris", booster = "dart", rate_drop = 0.05)

set.seed(31)

Expand Down
2 changes: 1 addition & 1 deletion src/test/java/org/jpmml/xgboost/ClassificationTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ public ClassificationTest(){

@Test
public void evaluateBinomialAudit() throws Exception {
evaluate("BinomialClassification", "Audit", new FloatEquivalence(16));
evaluate("BinomialClassification", "Audit", new FloatEquivalence(32));
}

@Test
Expand Down
Loading

0 comments on commit 0b8dade

Please sign in to comment.