scratch/redTime/lookatRedTimeResults_ExpRanges2.R
In CollinErickson/CGGP: Composite Grid Gaussian Processes
# Using expandedRanges2 on redTime
x100 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n100_s0228_all_input.csv")[,-1]))
y100 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n100_s0228_all_output.csv")[,-1])))
x1000 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0303_all_input.csv")[,-1]))
y1000 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0303_all_output.csv")[,-1])))
x1000_2 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0304_all_input.csv")[,-1]))
y1000_2 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0304_all_output.csv")[,-1])))
# Ran SGGP on output dimension 50
rt.sggp.199 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-199.rds")
rt.sggp.299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-299.rds")
rt.sggp.399 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-399.rds")
rt.sggp.499 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-499.rds")
rt.sggp.599 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-599.rds")
rt.sggp.699 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-699.rds")
rt.sggp.799 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-799.rds")
rt.sggp.899 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-899.rds")
rt.sggp.999 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-999.rds")
rt.sggp.1099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1099.rds")
rt.sggp.1299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1299.rds")
rt.sggp.1699 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1699.rds")
rt.sggp.2099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-2099.rds")
rt.sggp.2499 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-2499.rds")
rt.sggp.3099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-3099.rds")
rt.sggp.4099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-4099.rds")
rt.sggp.5099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-5099.rds")
rt.sggp.6099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6099.rds")
rt.sggp.6299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6299.rds")
rt.sggp.6899 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6899.rds")
rt.sggp.7099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-7099.rds")
rt.sggp.8099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-8099.rds")
stats.rt.sggp.199 <- SGGPvalstats(rt.sggp.199, x1000, y1000[,50])
stats.rt.sggp.299 <- SGGPvalstats(rt.sggp.299, x1000, y1000[,50])
stats.rt.sggp.399 <- SGGPvalstats(rt.sggp.399, x1000, y1000[,50])
stats.rt.sggp.499 <- SGGPvalstats(rt.sggp.499, x1000, y1000[,50])
stats.rt.sggp.599 <- SGGPvalstats(rt.sggp.599, x1000, y1000[,50])
stats.rt.sggp.699 <- SGGPvalstats(rt.sggp.699, x1000, y1000[,50])
stats.rt.sggp.799 <- SGGPvalstats(rt.sggp.799, x1000, y1000[,50])
stats.rt.sggp.899 <- SGGPvalstats(rt.sggp.899, x1000, y1000[,50])
stats.rt.sggp.999 <- SGGPvalstats(rt.sggp.999, x1000, y1000[,50])
stats.rt.sggp.1099 <- SGGPvalstats(rt.sggp.1099, x1000, y1000[,50])
stats.rt.sggp.1299 <- SGGPvalstats(rt.sggp.1299, x1000, y1000[,50])
stats.rt.sggp.1699 <- SGGPvalstats(rt.sggp.1699, x1000, y1000[,50])
stats.rt.sggp.2099 <- SGGPvalstats(rt.sggp.2099, x1000, y1000[,50])
stats.rt.sggp.2499 <- SGGPvalstats(rt.sggp.2499, x1000, y1000[,50])
stats.rt.sggp.3099 <- SGGPvalstats(rt.sggp.3099, x1000, y1000[,50])
stats.rt.sggp.4099 <- SGGPvalstats(rt.sggp.4099, x1000, y1000[,50])
stats.rt.sggp.5099 <- SGGPvalstats(rt.sggp.5099, x1000, y1000[,50])
stats.rt.sggp.6099 <- SGGPvalstats(rt.sggp.6099, x1000, y1000[,50])
stats.rt.sggp.6299 <- SGGPvalstats(rt.sggp.6299, x1000, y1000[,50])
stats.rt.sggp.6899 <- SGGPvalstats(rt.sggp.6899, x1000, y1000[,50])
stats.rt.sggp.7099 <- SGGPvalstats(rt.sggp.7099, x1000, y1000[,50])
stats.rt.sggp.8099 <- SGGPvalstats(rt.sggp.8099, x1000, y1000[,50])
# SGGP with other correlation functions
stats.rt.sggp.199.m3 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.199.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.199.pe <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.599.m3 <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.599.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.599.pe <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.1299.m3 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.1299.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.1299.pe <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.2499.m3 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.2499.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.2499.pe <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.6099.m3 <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.6099.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.6099.pe <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="pe"), x1000, y1000[,50])
# SGGP with more supp data
stats.rt.sggp.199.300 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.199.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.199.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.199.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.699.300 <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.699.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.699.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.699.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.1299.300 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.1299.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.1299.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.1299.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.2499.300 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.2499.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.2499.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.2499.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
# Run with mlegp
mod.mlegp.50 <- mlegp::mlegp(x100[1:50,], y100[1:50,50])
pred.mlegp.50 <- predict(mod.mlegp.50, x1000, se=T)
stats.mlegp.50 <- valstats(pred.mlegp.50$fit, pred.mlegp.50$se, y1000[,50])
mod.mlegp.75 <- mlegp::mlegp(x100[1:75,], y100[1:75,50])
pred.mlegp.75 <- predict(mod.mlegp.75, x1000, se=T)
stats.mlegp.75 <- valstats(pred.mlegp.75$fit, pred.mlegp.75$se, y1000[,50])
mod.mlegp.100 <- mlegp::mlegp(x100, y100[,50])
pred.mlegp.100 <- predict(mod.mlegp.100, x1000, se=T)
stats.mlegp.100 <- valstats(pred.mlegp.100$fit, pred.mlegp.100$se, y1000[,50])
mod.mlegp.200 <- mlegp::mlegp(x1000_2[1:200,], y1000_2[1:200,50])
pred.mlegp.200 <- predict(mod.mlegp.200, x1000, se=T)
stats.mlegp.200 <- valstats(pred.mlegp.200$fit, pred.mlegp.200$se, y1000[,50])
mod.mlegp.300 <- mlegp::mlegp(x1000_2[1:300,], y1000_2[1:300,50])
pred.mlegp.300 <- predict(mod.mlegp.300, x1000, se=T)
stats.mlegp.300 <- valstats(pred.mlegp.300$fit, pred.mlegp.300$se, y1000[,50])
mod.mlegp.400 <- mlegp::mlegp(x1000_2[1:400,], y1000_2[1:400,50])
pred.mlegp.400 <- predict(mod.mlegp.400, x1000, se=T)
stats.mlegp.400 <- valstats(pred.mlegp.400$fit, pred.mlegp.400$se, y1000[,50])
mod.mlegp.500 <- mlegp::mlegp(x1000_2[1:500,], y1000_2[1:500,50])
pred.mlegp.500 <- predict(mod.mlegp.500, x1000, se=T)
stats.mlegp.500 <- valstats(pred.mlegp.500$fit, pred.mlegp.500$se, y1000[,50])
# Run with DK
mod.DK.100 <- DiceKriging::km(design=x100, response=y100[,50])
pred.DK.100 <- DiceKriging::predict.km(mod.DK.100, x1000, se=T, type = "SK")
stats.DK.100 <- valstats(pred.DK.100$mean, pred.DK.100$sd, y1000[,50])
mod.DK.200 <- DiceKriging::km(design=x1000_2[1:200,], response=y1000_2[1:200,50])
pred.DK.200 <- DiceKriging::predict.km(mod.DK.200, x1000, se=T, type = "SK")
stats.DK.200 <- valstats(pred.DK.200$mean, pred.DK.200$sd, y1000[,50])
mod.DK.300 <- DiceKriging::km(design=x1000_2[1:300,], response=y1000_2[1:300,50])
pred.DK.300 <- DiceKriging::predict.km(mod.DK.300, x1000, se=T, type = "SK")
stats.DK.300 <- valstats(pred.DK.300$mean, pred.DK.300$sd, y1000[,50])
mod.DK.400 <- DiceKriging::km(design=x1000_2[1:400,], response=y1000_2[1:400,50])
pred.DK.400 <- DiceKriging::predict.km(mod.DK.400, x1000, se=T, type = "SK")
stats.DK.400 <- valstats(pred.DK.400$mean, pred.DK.400$sd, y1000[,50])
mod.DK.500 <- DiceKriging::km(design=x1000_2[1:500,], response=y1000_2[1:500,50])
pred.DK.500 <- DiceKriging::predict.km(mod.DK.500, x1000, se=T, type = "SK")
stats.DK.500 <- valstats(pred.DK.500$mean, pred.DK.500$sd, y1000[,50])
# 100+199 0.03076381 -6.262226 0.01522756 0.927 0.9996693 0.9993362
stats50 <- list(
# SGGP as fit while running redTime
data.frame("SGGP", 100, 199, 0.03076381, -6.262226, 0.01522756, 0.927, 0.9996693, 0.9993362, "CauchySQ")
,data.frame("SGGP", 100, 299, 0.03451267, -5.969878, 0.01650027, 0.904, 0.9995829, 0.9991646, "CauchySQ")
,data.frame("SGGP", 100, 399, 0.03300229, -6.088676, 0.01463064, 0.913, 0.9996187, 0.9992361, "CauchySQ")
,data.frame("SGGP", 100, 499, 0.02625919, -6.672529, 0.0116401, 0.945, 0.9997592, 0.9995164, "CauchySQ")
,data.frame("SGGP", 100, 599, 0.02180873, -6.890593, 0.01057118 , 0.951, 0.9998343, 0.9996664, "CauchySQ")
,data.frame("SGGP", 100, 699, 0.01777888, -7.284645, 0.008713988, 0.945, 0.9998899, 0.9997783, "CauchySQ")
,data.frame("SGGP", 100, 799, 0.02116162, -7.101041, 0.009290412, 0.946, 0.999843 , 0.9996859, "CauchySQ")
,data.frame("SGGP", 100, 899, 0.02314205, -6.917019, 0.009970185, 0.959, 0.9998133, 0.9996244, "CauchySQ")
,data.frame("SGGP", 100, 999, 0.02157915, -7.034552, 0.009963331, 0.947, 0.9998376, 0.9996734, "CauchySQ")
,data.frame("SGGP", 100, 1099, 0.01534432, -7.531176, 0.007727928, 0.969, 0.999918 , 0.9998349, "CauchySQ")
,data.frame("SGGP", 100, 1299, 0.01276961, -7.843283, 0.00664756 , 0.97 , 0.9999435, 0.9998856, "CauchySQ")
,data.frame("SGGP", 100, 1699, 0.01244242, -7.857461, 0.006426012, 0.97 , 0.9999461, 0.9998914, "CauchySQ")
,data.frame("SGGP", 100, 2099, 0.01097685, -8.078817, 0.005756391, 0.964, 0.9999579, 0.9999155, "CauchySQ")
,data.frame("SGGP", 100, 2499, 0.0100807, -8.237784, 0.005289297, 0.966, 0.9999644, 0.9999287, "CauchySQ")
,data.frame("SGGP", 100, 3099, 0.008719812, -8.298772, 0.004743257, 0.981, 0.9999734, 0.9999467, "CauchySQ")
,data.frame("SGGP", 100, 4099, 0.006635083, -8.847274, 0.003542177, 0.985, 0.9999846, 0.9999691, "CauchySQ")
,data.frame("SGGP", 100, 5099, 0.005979838, -9.149594, 0.003096542, 0.976, 0.9999875, 0.9999749, "CauchySQ")
,data.frame("SGGP", 100, 6099, 0.004925651, -9.422376, 0.002610824, 0.983, 0.9999915, 0.999983, "CauchySQ")
,data.frame("SGGP", 100, 6299, 0.004897624, -9.426733, 0.002620672, 0.986, 0.9999916, 0.9999832, "CauchySQ")
,data.frame("SGGP", 100, 6899, 0.004145463, -9.55806, 0.002485635, 0.982, 0.999994, 0.9999879, "CauchySQ")
,data.frame("SGGP", 100, 7099, 0.004176226, -9.597928, 0.002454114, 0.982, 0.9999939, 0.9999878, "CauchySQ")
,data.frame("SGGP", 100, 8099, 0.003917667, -9.773086, 0.002275711, 0.982, 0.9999946, 0.9999892, "CauchySQ")
# Refit with other correlation functions
,data.frame("SGGP", 100, 199, 0.03216928, -6.083981, 0.01645107, 0.981, 0.9996371, 0.9992742, "m32")
,data.frame("SGGP", 100, 199, 0.03084337, -6.275035, 0.01536157, 0.949, 0.9996666, 0.9993328, "Cauchy")
,data.frame("SGGP", 100, 199, 0.03434129, -5.302393, 0.02072374, 0.995, 0.9995864, 0.9991729, "PowExp")
,data.frame("SGGP", 100, 599, 0.02210409, -6.564741, 0.01173468, 0.984, 0.9998289, 0.9996573, "m32")
,data.frame("SGGP", 100, 599, 0.02388705, -6.535353, 0.01225433, 0.964, 0.999801 , 0.9995998, "Cauchy")
,data.frame("SGGP", 100, 599, 0.02642953, -5.05874, 0.02094761, 0.998, 0.9997551, 0.9995101, "PowExp")
,data.frame("SGGP", 100, 1299, 0.01294041, -6.807223, 0.009246659, 1, 0.9999425, 0.9998826, "m32")
,data.frame("SGGP", 100, 1299, 0.01329422, -7.734564, 0.007000892, 0.971, 0.9999392, 0.999876, "Cauchy")
,data.frame("SGGP", 100, 1299, 0.01482763, -6.200806, 0.01194623, 1, 0.9999259, 0.9998458, "PowExp")
,data.frame("SGGP", 100, 2499, 0.01060865, -7.238947, 0.007474609, 0.999, 0.9999607, 0.9999211, "m32")
,data.frame("SGGP", 100, 2499, 0.01032681, -8.140031, 0.005506091, 0.97, 0.9999627, 0.9999252, "Cauchy")
,data.frame("SGGP", 100, 2499, 0.01154785, -6.110847, 0.01194843, 1, 0.9999535, 0.9999065, "PowExp")
,data.frame("SGGP", 100, 6099, 0.006155724, -7.200618, 0.0070135, 0.999, 0.9999868, 0.9999734, "m32")
,data.frame("SGGP", 100, 6099, 0.004818425, -8.291389, 0.00414775, 0.997, 0.9999919, 0.9999837, "Cauchy")
,data.frame("SGGP", 100, 6099, 0.004711587, -7.111454, 0.006963024, 1, 0.9999922, 0.9999844, "PowExp")
# Refit with 300 supp pts
,data.frame("SGGP", 300, 199, 0.02247822, -7.482398, 0.00928634, 0.92, 0.9998228, 0.9996456, "CauchySQ")
,data.frame("SGGP", 300, 199, 0.02451253, -7.288525, 0.01043526, 0.953, 0.9997895, 0.9995786, "m32")
,data.frame("SGGP", 300, 199, 0.02371664, -7.288193, 0.009969655, 0.931, 0.9998029, 0.9996055, "Cauchy")
,data.frame("SGGP", 300, 199, 0.02927608, -6.341542, 0.01419704, 0.986, 0.9997014, 0.9993989, "PowExp")
,data.frame("SGGP", 300, 699, 0.01802544, -7.68001, 0.007389155, 0.934, 0.9998865, 0.9997721, "CauchySQ")
,data.frame("SGGP", 300, 699, 0.01808537, -7.52286, 0.008136096, 0.977, 0.9998864, 0.9997706, "m32")
,data.frame("SGGP", 300, 699, 0.01974602, -7.42624, 0.008304455, 0.94, 0.9998644, 0.9997265, "Cauchy")
,data.frame("SGGP", 300, 699, 0.02370324, -5.997354, 0.01435612 , 0.993, 0.9998032, 0.9996059, "PowExp")
,data.frame("SGGP", 300, 1299, 0.0117054, -8.342819, 0.005552262, 0.964, 0.999952, 0.9999039, "CauchySQ")
,data.frame("SGGP", 300, 1299, 0.01226761, -7.426326, 0.007429247, 0.999, 0.9999476, 0.9998944, "m32")
,data.frame("SGGP", 300, 1299, 0.0116426, -8.366771, 0.005529094, 0.97, 0.9999526, 0.9999049, "Cauchy")
,data.frame("SGGP", 300, 1299, 0.01342289, -6.753954, 0.009527712, 0.999, 0.9999379, 0.9998736, "PowExp")
,data.frame("SGGP", 300, 2499, 0.009238336, -8.688343, 0.004462736, 0.956, 0.9999702, 0.9999401, "CauchySQ")
,data.frame("SGGP", 300, 2499, 0.009186548, -7.760675, 0.005993132, 0.998, 0.9999706, 0.9999408, "m32")
,data.frame("SGGP", 300, 2499, 0.009156846, -8.667511, 0.00446411 , 0.974, 0.9999707, 0.9999412, "Cauchy")
,data.frame("SGGP", 300, 2499, 0.01136375 , -5.969252, 0.01314411 , 1, 0.9999579, 0.9999094, "PowExp")
# ,data.frame("SGGP", 100, 199, )
# ,data.frame("SGGP", 100, 199, )
,data.frame("mlegp", 0,50, 0.2270544, -1.913619, 0.1250952, 0.999, 0.9831718, 0.9638425, "gauss")
,data.frame("mlegp", 0,75, 0.10766, -2.792874, 0.07135557, 1, 0.996209, 0.9918708, "gauss")
,data.frame("mlegp", 0,100, 0.06409675, -3.36158, 0.05061328, 1, 0.9985716, 0.9971186, "gauss")
,data.frame("mlegp", 0,200, 0.06407573, 157685457, 0.03691848, 0.986, 0.9985597, 0.9971204, "gauss")
,data.frame("mlegp", 0,300, 0.05759601, 1169074, 0.03370455, 0.998, 0.9988548, 0.9976734, "gauss")
,data.frame("mlegp", 0,400, 0.05136801, -4.237131, 0.03544184, 1, 0.9990808, 0.9981494, "gauss")
# ,data.frame("mlegp", 0, 100, )
# ,data.frame("mlegp", 0, 100, )
,data.frame("DK", 0,100, 0.1124202, -2.070725, 0.09362545, 1, 0.9958498, 0.9911361, "m52")
,data.frame("DK", 0,200, 0.154419, -2.548706, 0.08653054, 0.997, 0.995057, 0.983276, "m52")
,data.frame("DK", 0,300, 0.1016811, -3.085017, 0.06229843, 0.996, 0.9974675, 0.9927487, "m52")
,data.frame("DK", 0,400, 0.07295021, -3.397965, 0.05084728, 0.997, 0.9984053, 0.9962676, "m52")
,data.frame("DK", 0,500, 0.05559626, -3.598408, 0.04458003, 1, 0.9989885, 0.9978321, "m52")
)
stats50 <- lapply(stats50, function(x){colnames(x) <- c("Package", 'Nsup',"Ngrid","RMSE","score","CRPscore","coverage","corr","R2","Corr");x})
stats50 <- do.call(rbind, stats50)
stats50$Ntotal <- stats50$Nsup + stats50$Ngrid
library(ggplot2)
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3)
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3) + scale_x_log10() + scale_y_log10()
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=Nsup, shape=Corr)) + geom_point(size=3) + facet_grid(. ~ Package) + scale_x_log10() + scale_y_log10()
ggplot(data=stats50 %>% filter(score<1e5), mapping=aes(Ntotal, score, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3) + scale_x_log10()
CollinErickson/CGGP documentation built on Feb. 6, 2024, 2:24 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# Using expandedRanges2 on redTime
x100 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n100_s0228_all_input.csv")[,-1]))
y100 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n100_s0228_all_output.csv")[,-1])))
x1000 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0303_all_input.csv")[,-1]))
y1000 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0303_all_output.csv")[,-1])))
x1000_2 <- unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0304_all_input.csv")[,-1]))
y1000_2 <- log(unname(as.matrix(read.csv("./scratch/redTime/redTimeData/ExpandedRanges2_LHS1L_n1000_s0304_all_output.csv")[,-1])))
# Ran SGGP on output dimension 50
rt.sggp.199 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-199.rds")
rt.sggp.299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-299.rds")
rt.sggp.399 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-399.rds")
rt.sggp.499 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-499.rds")
rt.sggp.599 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-599.rds")
rt.sggp.699 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-699.rds")
rt.sggp.799 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-799.rds")
rt.sggp.899 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-899.rds")
rt.sggp.999 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-999.rds")
rt.sggp.1099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1099.rds")
rt.sggp.1299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1299.rds")
rt.sggp.1699 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-1699.rds")
rt.sggp.2099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-2099.rds")
rt.sggp.2499 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-2499.rds")
rt.sggp.3099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-3099.rds")
rt.sggp.4099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-4099.rds")
rt.sggp.5099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-5099.rds")
rt.sggp.6099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6099.rds")
rt.sggp.6299 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6299.rds")
rt.sggp.6899 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-6899.rds")
rt.sggp.7099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-7099.rds")
rt.sggp.8099 <- readRDS("./scratch/redTime/redTimeData/out_S2o50_SGGP-8099.rds")
stats.rt.sggp.199 <- SGGPvalstats(rt.sggp.199, x1000, y1000[,50])
stats.rt.sggp.299 <- SGGPvalstats(rt.sggp.299, x1000, y1000[,50])
stats.rt.sggp.399 <- SGGPvalstats(rt.sggp.399, x1000, y1000[,50])
stats.rt.sggp.499 <- SGGPvalstats(rt.sggp.499, x1000, y1000[,50])
stats.rt.sggp.599 <- SGGPvalstats(rt.sggp.599, x1000, y1000[,50])
stats.rt.sggp.699 <- SGGPvalstats(rt.sggp.699, x1000, y1000[,50])
stats.rt.sggp.799 <- SGGPvalstats(rt.sggp.799, x1000, y1000[,50])
stats.rt.sggp.899 <- SGGPvalstats(rt.sggp.899, x1000, y1000[,50])
stats.rt.sggp.999 <- SGGPvalstats(rt.sggp.999, x1000, y1000[,50])
stats.rt.sggp.1099 <- SGGPvalstats(rt.sggp.1099, x1000, y1000[,50])
stats.rt.sggp.1299 <- SGGPvalstats(rt.sggp.1299, x1000, y1000[,50])
stats.rt.sggp.1699 <- SGGPvalstats(rt.sggp.1699, x1000, y1000[,50])
stats.rt.sggp.2099 <- SGGPvalstats(rt.sggp.2099, x1000, y1000[,50])
stats.rt.sggp.2499 <- SGGPvalstats(rt.sggp.2499, x1000, y1000[,50])
stats.rt.sggp.3099 <- SGGPvalstats(rt.sggp.3099, x1000, y1000[,50])
stats.rt.sggp.4099 <- SGGPvalstats(rt.sggp.4099, x1000, y1000[,50])
stats.rt.sggp.5099 <- SGGPvalstats(rt.sggp.5099, x1000, y1000[,50])
stats.rt.sggp.6099 <- SGGPvalstats(rt.sggp.6099, x1000, y1000[,50])
stats.rt.sggp.6299 <- SGGPvalstats(rt.sggp.6299, x1000, y1000[,50])
stats.rt.sggp.6899 <- SGGPvalstats(rt.sggp.6899, x1000, y1000[,50])
stats.rt.sggp.7099 <- SGGPvalstats(rt.sggp.7099, x1000, y1000[,50])
stats.rt.sggp.8099 <- SGGPvalstats(rt.sggp.8099, x1000, y1000[,50])
# SGGP with other correlation functions
stats.rt.sggp.199.m3 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.199.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.199.pe <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=rt.sggp.199$Xs, Ys=rt.sggp.199$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.599.m3 <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.599.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.599.pe <- SGGPvalstats(SGGPfit(rt.sggp.599, rt.sggp.599$Y, Xs=rt.sggp.599$Xs, Ys=rt.sggp.599$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.1299.m3 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.1299.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.1299.pe <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=rt.sggp.1299$Xs, Ys=rt.sggp.1299$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.2499.m3 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.2499.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.2499.pe <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=rt.sggp.2499$Xs, Ys=rt.sggp.2499$Ys, corr="pe"), x1000, y1000[,50])
stats.rt.sggp.6099.m3 <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="m32"), x1000, y1000[,50])
stats.rt.sggp.6099.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.6099.pe <- SGGPvalstats(SGGPfit(rt.sggp.6099, rt.sggp.6099$Y, Xs=rt.sggp.6099$Xs, Ys=rt.sggp.6099$Ys, corr="pe"), x1000, y1000[,50])
# SGGP with more supp data
stats.rt.sggp.199.300 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.199.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.199.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.199.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.199, rt.sggp.199$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.699.300 <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.699.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.699.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.699.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.699, rt.sggp.699$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.1299.300 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.1299.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.1299.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.1299.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.1299, rt.sggp.1299$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
stats.rt.sggp.2499.300 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50]), x1000, y1000[,50])
stats.rt.sggp.2499.300.m3 <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="m32"), x1000, y1000[,50])
stats.rt.sggp.2499.300.cauchy <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="cauchy"), x1000, y1000[,50])
stats.rt.sggp.2499.300.pe <- SGGPvalstats(SGGPfit(rt.sggp.2499, rt.sggp.2499$Y, Xs=x1000_2[1:300,], Ys=y1000_2[1:300,50], corr="pe"), x1000, y1000[,50])
# Run with mlegp
mod.mlegp.50 <- mlegp::mlegp(x100[1:50,], y100[1:50,50])
pred.mlegp.50 <- predict(mod.mlegp.50, x1000, se=T)
stats.mlegp.50 <- valstats(pred.mlegp.50$fit, pred.mlegp.50$se, y1000[,50])
mod.mlegp.75 <- mlegp::mlegp(x100[1:75,], y100[1:75,50])
pred.mlegp.75 <- predict(mod.mlegp.75, x1000, se=T)
stats.mlegp.75 <- valstats(pred.mlegp.75$fit, pred.mlegp.75$se, y1000[,50])
mod.mlegp.100 <- mlegp::mlegp(x100, y100[,50])
pred.mlegp.100 <- predict(mod.mlegp.100, x1000, se=T)
stats.mlegp.100 <- valstats(pred.mlegp.100$fit, pred.mlegp.100$se, y1000[,50])
mod.mlegp.200 <- mlegp::mlegp(x1000_2[1:200,], y1000_2[1:200,50])
pred.mlegp.200 <- predict(mod.mlegp.200, x1000, se=T)
stats.mlegp.200 <- valstats(pred.mlegp.200$fit, pred.mlegp.200$se, y1000[,50])
mod.mlegp.300 <- mlegp::mlegp(x1000_2[1:300,], y1000_2[1:300,50])
pred.mlegp.300 <- predict(mod.mlegp.300, x1000, se=T)
stats.mlegp.300 <- valstats(pred.mlegp.300$fit, pred.mlegp.300$se, y1000[,50])
mod.mlegp.400 <- mlegp::mlegp(x1000_2[1:400,], y1000_2[1:400,50])
pred.mlegp.400 <- predict(mod.mlegp.400, x1000, se=T)
stats.mlegp.400 <- valstats(pred.mlegp.400$fit, pred.mlegp.400$se, y1000[,50])
mod.mlegp.500 <- mlegp::mlegp(x1000_2[1:500,], y1000_2[1:500,50])
pred.mlegp.500 <- predict(mod.mlegp.500, x1000, se=T)
stats.mlegp.500 <- valstats(pred.mlegp.500$fit, pred.mlegp.500$se, y1000[,50])
# Run with DK
mod.DK.100 <- DiceKriging::km(design=x100, response=y100[,50])
pred.DK.100 <- DiceKriging::predict.km(mod.DK.100, x1000, se=T, type = "SK")
stats.DK.100 <- valstats(pred.DK.100$mean, pred.DK.100$sd, y1000[,50])
mod.DK.200 <- DiceKriging::km(design=x1000_2[1:200,], response=y1000_2[1:200,50])
pred.DK.200 <- DiceKriging::predict.km(mod.DK.200, x1000, se=T, type = "SK")
stats.DK.200 <- valstats(pred.DK.200$mean, pred.DK.200$sd, y1000[,50])
mod.DK.300 <- DiceKriging::km(design=x1000_2[1:300,], response=y1000_2[1:300,50])
pred.DK.300 <- DiceKriging::predict.km(mod.DK.300, x1000, se=T, type = "SK")
stats.DK.300 <- valstats(pred.DK.300$mean, pred.DK.300$sd, y1000[,50])
mod.DK.400 <- DiceKriging::km(design=x1000_2[1:400,], response=y1000_2[1:400,50])
pred.DK.400 <- DiceKriging::predict.km(mod.DK.400, x1000, se=T, type = "SK")
stats.DK.400 <- valstats(pred.DK.400$mean, pred.DK.400$sd, y1000[,50])
mod.DK.500 <- DiceKriging::km(design=x1000_2[1:500,], response=y1000_2[1:500,50])
pred.DK.500 <- DiceKriging::predict.km(mod.DK.500, x1000, se=T, type = "SK")
stats.DK.500 <- valstats(pred.DK.500$mean, pred.DK.500$sd, y1000[,50])
# 100+199 0.03076381 -6.262226 0.01522756 0.927 0.9996693 0.9993362
stats50 <- list(
# SGGP as fit while running redTime
data.frame("SGGP", 100, 199, 0.03076381, -6.262226, 0.01522756, 0.927, 0.9996693, 0.9993362, "CauchySQ")
,data.frame("SGGP", 100, 299, 0.03451267, -5.969878, 0.01650027, 0.904, 0.9995829, 0.9991646, "CauchySQ")
,data.frame("SGGP", 100, 399, 0.03300229, -6.088676, 0.01463064, 0.913, 0.9996187, 0.9992361, "CauchySQ")
,data.frame("SGGP", 100, 499, 0.02625919, -6.672529, 0.0116401, 0.945, 0.9997592, 0.9995164, "CauchySQ")
,data.frame("SGGP", 100, 599, 0.02180873, -6.890593, 0.01057118 , 0.951, 0.9998343, 0.9996664, "CauchySQ")
,data.frame("SGGP", 100, 699, 0.01777888, -7.284645, 0.008713988, 0.945, 0.9998899, 0.9997783, "CauchySQ")
,data.frame("SGGP", 100, 799, 0.02116162, -7.101041, 0.009290412, 0.946, 0.999843 , 0.9996859, "CauchySQ")
,data.frame("SGGP", 100, 899, 0.02314205, -6.917019, 0.009970185, 0.959, 0.9998133, 0.9996244, "CauchySQ")
,data.frame("SGGP", 100, 999, 0.02157915, -7.034552, 0.009963331, 0.947, 0.9998376, 0.9996734, "CauchySQ")
,data.frame("SGGP", 100, 1099, 0.01534432, -7.531176, 0.007727928, 0.969, 0.999918 , 0.9998349, "CauchySQ")
,data.frame("SGGP", 100, 1299, 0.01276961, -7.843283, 0.00664756 , 0.97 , 0.9999435, 0.9998856, "CauchySQ")
,data.frame("SGGP", 100, 1699, 0.01244242, -7.857461, 0.006426012, 0.97 , 0.9999461, 0.9998914, "CauchySQ")
,data.frame("SGGP", 100, 2099, 0.01097685, -8.078817, 0.005756391, 0.964, 0.9999579, 0.9999155, "CauchySQ")
,data.frame("SGGP", 100, 2499, 0.0100807, -8.237784, 0.005289297, 0.966, 0.9999644, 0.9999287, "CauchySQ")
,data.frame("SGGP", 100, 3099, 0.008719812, -8.298772, 0.004743257, 0.981, 0.9999734, 0.9999467, "CauchySQ")
,data.frame("SGGP", 100, 4099, 0.006635083, -8.847274, 0.003542177, 0.985, 0.9999846, 0.9999691, "CauchySQ")
,data.frame("SGGP", 100, 5099, 0.005979838, -9.149594, 0.003096542, 0.976, 0.9999875, 0.9999749, "CauchySQ")
,data.frame("SGGP", 100, 6099, 0.004925651, -9.422376, 0.002610824, 0.983, 0.9999915, 0.999983, "CauchySQ")
,data.frame("SGGP", 100, 6299, 0.004897624, -9.426733, 0.002620672, 0.986, 0.9999916, 0.9999832, "CauchySQ")
,data.frame("SGGP", 100, 6899, 0.004145463, -9.55806, 0.002485635, 0.982, 0.999994, 0.9999879, "CauchySQ")
,data.frame("SGGP", 100, 7099, 0.004176226, -9.597928, 0.002454114, 0.982, 0.9999939, 0.9999878, "CauchySQ")
,data.frame("SGGP", 100, 8099, 0.003917667, -9.773086, 0.002275711, 0.982, 0.9999946, 0.9999892, "CauchySQ")
# Refit with other correlation functions
,data.frame("SGGP", 100, 199, 0.03216928, -6.083981, 0.01645107, 0.981, 0.9996371, 0.9992742, "m32")
,data.frame("SGGP", 100, 199, 0.03084337, -6.275035, 0.01536157, 0.949, 0.9996666, 0.9993328, "Cauchy")
,data.frame("SGGP", 100, 199, 0.03434129, -5.302393, 0.02072374, 0.995, 0.9995864, 0.9991729, "PowExp")
,data.frame("SGGP", 100, 599, 0.02210409, -6.564741, 0.01173468, 0.984, 0.9998289, 0.9996573, "m32")
,data.frame("SGGP", 100, 599, 0.02388705, -6.535353, 0.01225433, 0.964, 0.999801 , 0.9995998, "Cauchy")
,data.frame("SGGP", 100, 599, 0.02642953, -5.05874, 0.02094761, 0.998, 0.9997551, 0.9995101, "PowExp")
,data.frame("SGGP", 100, 1299, 0.01294041, -6.807223, 0.009246659, 1, 0.9999425, 0.9998826, "m32")
,data.frame("SGGP", 100, 1299, 0.01329422, -7.734564, 0.007000892, 0.971, 0.9999392, 0.999876, "Cauchy")
,data.frame("SGGP", 100, 1299, 0.01482763, -6.200806, 0.01194623, 1, 0.9999259, 0.9998458, "PowExp")
,data.frame("SGGP", 100, 2499, 0.01060865, -7.238947, 0.007474609, 0.999, 0.9999607, 0.9999211, "m32")
,data.frame("SGGP", 100, 2499, 0.01032681, -8.140031, 0.005506091, 0.97, 0.9999627, 0.9999252, "Cauchy")
,data.frame("SGGP", 100, 2499, 0.01154785, -6.110847, 0.01194843, 1, 0.9999535, 0.9999065, "PowExp")
,data.frame("SGGP", 100, 6099, 0.006155724, -7.200618, 0.0070135, 0.999, 0.9999868, 0.9999734, "m32")
,data.frame("SGGP", 100, 6099, 0.004818425, -8.291389, 0.00414775, 0.997, 0.9999919, 0.9999837, "Cauchy")
,data.frame("SGGP", 100, 6099, 0.004711587, -7.111454, 0.006963024, 1, 0.9999922, 0.9999844, "PowExp")
# Refit with 300 supp pts
,data.frame("SGGP", 300, 199, 0.02247822, -7.482398, 0.00928634, 0.92, 0.9998228, 0.9996456, "CauchySQ")
,data.frame("SGGP", 300, 199, 0.02451253, -7.288525, 0.01043526, 0.953, 0.9997895, 0.9995786, "m32")
,data.frame("SGGP", 300, 199, 0.02371664, -7.288193, 0.009969655, 0.931, 0.9998029, 0.9996055, "Cauchy")
,data.frame("SGGP", 300, 199, 0.02927608, -6.341542, 0.01419704, 0.986, 0.9997014, 0.9993989, "PowExp")
,data.frame("SGGP", 300, 699, 0.01802544, -7.68001, 0.007389155, 0.934, 0.9998865, 0.9997721, "CauchySQ")
,data.frame("SGGP", 300, 699, 0.01808537, -7.52286, 0.008136096, 0.977, 0.9998864, 0.9997706, "m32")
,data.frame("SGGP", 300, 699, 0.01974602, -7.42624, 0.008304455, 0.94, 0.9998644, 0.9997265, "Cauchy")
,data.frame("SGGP", 300, 699, 0.02370324, -5.997354, 0.01435612 , 0.993, 0.9998032, 0.9996059, "PowExp")
,data.frame("SGGP", 300, 1299, 0.0117054, -8.342819, 0.005552262, 0.964, 0.999952, 0.9999039, "CauchySQ")
,data.frame("SGGP", 300, 1299, 0.01226761, -7.426326, 0.007429247, 0.999, 0.9999476, 0.9998944, "m32")
,data.frame("SGGP", 300, 1299, 0.0116426, -8.366771, 0.005529094, 0.97, 0.9999526, 0.9999049, "Cauchy")
,data.frame("SGGP", 300, 1299, 0.01342289, -6.753954, 0.009527712, 0.999, 0.9999379, 0.9998736, "PowExp")
,data.frame("SGGP", 300, 2499, 0.009238336, -8.688343, 0.004462736, 0.956, 0.9999702, 0.9999401, "CauchySQ")
,data.frame("SGGP", 300, 2499, 0.009186548, -7.760675, 0.005993132, 0.998, 0.9999706, 0.9999408, "m32")
,data.frame("SGGP", 300, 2499, 0.009156846, -8.667511, 0.00446411 , 0.974, 0.9999707, 0.9999412, "Cauchy")
,data.frame("SGGP", 300, 2499, 0.01136375 , -5.969252, 0.01314411 , 1, 0.9999579, 0.9999094, "PowExp")
# ,data.frame("SGGP", 100, 199, )
# ,data.frame("SGGP", 100, 199, )
,data.frame("mlegp", 0,50, 0.2270544, -1.913619, 0.1250952, 0.999, 0.9831718, 0.9638425, "gauss")
,data.frame("mlegp", 0,75, 0.10766, -2.792874, 0.07135557, 1, 0.996209, 0.9918708, "gauss")
,data.frame("mlegp", 0,100, 0.06409675, -3.36158, 0.05061328, 1, 0.9985716, 0.9971186, "gauss")
,data.frame("mlegp", 0,200, 0.06407573, 157685457, 0.03691848, 0.986, 0.9985597, 0.9971204, "gauss")
,data.frame("mlegp", 0,300, 0.05759601, 1169074, 0.03370455, 0.998, 0.9988548, 0.9976734, "gauss")
,data.frame("mlegp", 0,400, 0.05136801, -4.237131, 0.03544184, 1, 0.9990808, 0.9981494, "gauss")
# ,data.frame("mlegp", 0, 100, )
# ,data.frame("mlegp", 0, 100, )
,data.frame("DK", 0,100, 0.1124202, -2.070725, 0.09362545, 1, 0.9958498, 0.9911361, "m52")
,data.frame("DK", 0,200, 0.154419, -2.548706, 0.08653054, 0.997, 0.995057, 0.983276, "m52")
,data.frame("DK", 0,300, 0.1016811, -3.085017, 0.06229843, 0.996, 0.9974675, 0.9927487, "m52")
,data.frame("DK", 0,400, 0.07295021, -3.397965, 0.05084728, 0.997, 0.9984053, 0.9962676, "m52")
,data.frame("DK", 0,500, 0.05559626, -3.598408, 0.04458003, 1, 0.9989885, 0.9978321, "m52")
)
stats50 <- lapply(stats50, function(x){colnames(x) <- c("Package", 'Nsup',"Ngrid","RMSE","score","CRPscore","coverage","corr","R2","Corr");x})
stats50 <- do.call(rbind, stats50)
stats50$Ntotal <- stats50$Nsup + stats50$Ngrid
library(ggplot2)
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3)
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3) + scale_x_log10() + scale_y_log10()
ggplot(data=stats50, mapping=aes(Ntotal, RMSE, color=Nsup, shape=Corr)) + geom_point(size=3) + facet_grid(. ~ Package) + scale_x_log10() + scale_y_log10()
ggplot(data=stats50 %>% filter(score<1e5), mapping=aes(Ntotal, score, color=interaction(Package,Corr), shape=as.factor(Nsup))) + geom_point(size=3) + scale_x_log10()
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.