296 lines (295 with data), 13.8 kB
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[1] \"../data/simulations_20210329212141/equal/5/JI.txt\"\n",
" AE_FAETC X.AE_FCTAE X.DAE_FAETC X.DAE_FCTAE X.VAE_FCTAE X.LSTMVAE_FCTAE\n",
"1 0.7864754 0.8555881 0.7650096 0.8729629 0.5151694 0.5910766\n",
"2 0.7930283 0.8557713 0.7663618 0.8816597 0.5152971 0.5982880\n",
"3 0.7815983 0.8519293 0.7703164 0.8696588 0.5112708 0.5954603\n",
"4 0.7874296 0.8380773 0.7777636 0.8815903 0.5303783 0.6026711\n",
"5 0.7998520 0.8593818 0.7718466 0.8699363 0.5228915 0.5959956\n",
" [1] \"AE_FAETC_HET\" \"AE_FAETC_EQ\" \"X.AE_FCTAE_HET\" \n",
" [4] \"X.AE_FCTAE_EQ\" \"X.DAE_FAETC_HET\" \"X.DAE_FAETC_EQ\" \n",
" [7] \"X.DAE_FCTAE_HET\" \"X.DAE_FCTAE_EQ\" \"X.VAE_FCTAE_HET\" \n",
"[10] \"X.VAE_FCTAE_EQ\" \"X.LSTMVAE_FCTAE_HET\" \"X.LSTMVAE_FCTAE_EQ\" \n",
"[1] \"../data/simulations_20210329212141/equal/10/JI.txt\"\n",
" AE_FAETC X.AE_FCTAE X.DAE_FAETC X.DAE_FCTAE X.VAE_FCTAE X.LSTMVAE_FCTAE\n",
"1 0.7686000 0.7753077 0.7860100 0.8219333 0.7515000 0.4608212\n",
"2 0.7666667 0.7911071 0.7826930 0.8166000 0.7611857 0.4539229\n",
"3 0.7535500 0.7876333 0.7880110 0.8121667 0.7651143 0.4404087\n",
"4 0.7576000 0.7868333 0.7756981 0.8264000 0.7584012 0.4448950\n",
"5 0.7420833 0.7930950 0.7764907 0.8163333 0.7577496 0.4528135\n",
"6 0.7601000 0.7917452 0.7704681 0.8122667 0.7632258 0.4485148\n",
"7 0.7503833 0.7896333 0.7619731 0.8235000 0.7605238 0.4548279\n",
"8 0.7638190 0.7692190 0.7753311 0.8115667 0.7674000 0.4511056\n",
"9 0.7649667 0.7795667 0.7807456 0.8183000 0.7581286 0.4587046\n",
"10 0.7444429 0.7947190 0.7599756 0.8188286 0.7734583 0.4514359\n",
" [1] \"AE_FAETC_HET\" \"AE_FAETC_EQ\" \"X.AE_FCTAE_HET\" \n",
" [4] \"X.AE_FCTAE_EQ\" \"X.DAE_FAETC_HET\" \"X.DAE_FAETC_EQ\" \n",
" [7] \"X.DAE_FCTAE_HET\" \"X.DAE_FCTAE_EQ\" \"X.VAE_FCTAE_HET\" \n",
"[10] \"X.VAE_FCTAE_EQ\" \"X.LSTMVAE_FCTAE_HET\" \"X.LSTMVAE_FCTAE_EQ\" \n",
"[1] \"../data/simulations_20210329212141/equal/15/JI.txt\"\n",
" AE_FAETC X.AE_FCTAE X.DAE_FAETC X.DAE_FCTAE X.VAE_FCTAE X.LSTMVAE_FCTAE\n",
"1 0.7663878 0.7989517 0.7532231 0.7940330 0.7365600 0.4727026\n",
"2 0.7676835 0.8052115 0.7563619 0.8056647 0.7399893 0.4756525\n",
"3 0.7621207 0.8019024 0.7594553 0.7995865 0.7380597 0.4718643\n",
"4 0.7733462 0.8157678 0.7525584 0.7820629 0.7495614 0.4635589\n",
"5 0.7675834 0.7945885 0.7487293 0.7863840 0.7558608 0.4793765\n",
"6 0.7461658 0.8191150 0.7492135 0.7954261 0.7513054 0.4613873\n",
"7 0.7597422 0.8043023 0.7440258 0.7985005 0.7519942 0.4629963\n",
"8 0.7563764 0.8144615 0.7524720 0.7922929 0.7433741 0.4653092\n",
"9 0.7619438 0.8061806 0.7362969 0.7948400 0.7291959 0.4717958\n",
"10 0.7697440 0.8206591 0.7606681 0.8013052 0.7450379 0.4745232\n",
"11 0.7601137 0.8057421 0.7808652 0.8022179 0.7317692 0.4752839\n",
"12 0.7770244 0.7979146 0.7486006 0.7860330 0.7382499 0.4738429\n",
"13 0.7786453 0.8153197 0.7620518 0.8102876 0.7398166 0.4763239\n",
"14 0.7430794 0.8033851 0.7535825 0.7921890 0.7516091 0.4642153\n",
"15 0.7454310 0.8177235 0.7502729 0.7864014 0.7268284 0.4669012\n",
" [1] \"AE_FAETC_HET\" \"AE_FAETC_EQ\" \"X.AE_FCTAE_HET\" \n",
" [4] \"X.AE_FCTAE_EQ\" \"X.DAE_FAETC_HET\" \"X.DAE_FAETC_EQ\" \n",
" [7] \"X.DAE_FCTAE_HET\" \"X.DAE_FCTAE_EQ\" \"X.VAE_FCTAE_HET\" \n",
"[10] \"X.VAE_FCTAE_EQ\" \"X.LSTMVAE_FCTAE_HET\" \"X.LSTMVAE_FCTAE_EQ\" \n"
]
},
{
"data": {
"text/html": [
"<strong>png:</strong> 2"
],
"text/latex": [
"\\textbf{png:} 2"
],
"text/markdown": [
"**png:** 2"
],
"text/plain": [
"png \n",
" 2 "
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"list_clusters <- seq(5,15,5)\n",
"list_distrib <- c(\"heterogeneous\",\"equal\")\n",
"# Save all boxplots in a single PDF output file\n",
"results_folder<-\"../data/simulations_20210329212141/\"\n",
"pdf(file=paste0(results_folder, \"AE_simulated_boxplots.pdf\"), width = 15, height = 15, onefile = TRUE)\n",
"\n",
"# For each chosen number of clusters\n",
"for (i in list_clusters) {\n",
" \n",
" # Output files for each distribution\n",
" eq_file <- paste0(results_folder, \"equal/\", i,\"/JI.txt\")\n",
" het_file <- paste0(results_folder, \"heterogeneous/\",i, \"/JI.txt\")\n",
" print(eq_file)\n",
" if(exists(\"JI.final\")) rm(JI.final)\n",
" \n",
" # Load clusters (equal distribution)\n",
" if(file.exists(eq_file)) {\n",
" JI.final <- read.table(eq_file, sep=\"\\t\", header=TRUE)\n",
" print(JI.final)\n",
" names(JI.final) <- paste0(names(JI.final), \"_EQ\") \n",
" }\n",
" #Load clusters (heterogeneous distribution)\n",
" if(file.exists(het_file)) {\n",
" JI.het <- read.table(het_file, sep=\"\\t\", header=TRUE)\n",
" names_methods <- names(JI.het)\n",
" names(JI.het) <- paste0(names(JI.het), \"_HET\")\n",
"\n",
" # Aggregate results\n",
" if(exists(\"JI.final\")) {\n",
" JI.final <- data.frame(JI.het, JI.final)\n",
" new_order <- apply(expand.grid(c(\"_HET\", \"_EQ\"), names_methods)[, c(2,1)], 1, paste, collapse=\"\")\n",
" JI.final <- JI.final[, new_order]\n",
" }\n",
" else {\n",
" JI.final <- JI.het\n",
" }\n",
" }\n",
" print(names(JI.final))\n",
" \n",
" labels<-list(\"AE_FAETC_HET\",\"AE_FAETC_EQ\",\"AE_FCTAE_HET\",\"AE_FCTAE_EQ\",\"DAE_FAETC_HET\",\"DAE_FAETC_EQ\",\"DAE_FCTAE_HET\",\"DAE_FCTAE_EQ\",\"VAE_FCTAE_HET\" \n",
",\"VAE_FCTAE_EQ\",\"LSTMVAE_FCTAE_HET\",\"LSTMVAE_FCTAE_EQ\" )\n",
" # Plot results\n",
" par(mar=c(25,4,2,2)+.1)\n",
" boxplot(JI.final, xaxt=\"none\", cex.axis=3.5, \n",
" col=c('gray','gray','red','red','blue','blue','blueviolet','blueviolet','deeppink','deeppink','chocolate1','chocolate1'), \n",
" ann=FALSE, outline=FALSE)\n",
" matplot(1:ncol(JI.final), t(JI.final), col=\"black\", pch=16, xaxt=\"none\", cex=0.8, add=TRUE)\n",
" axis(1, at=1:ncol(JI.final), labels=labels, \n",
" las=2, srt=45, cex=0.5, cex.lab=2, cex.axis=2, cex.main=1.5, cex.sub=1.5) \n",
" title(main=paste(i,\"clusters\",sep=\" \"), \n",
" cex.lab=0.75, line = -2.5, adj=0, cex.main=3.5)\n",
"}\n",
"dev.off()"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<strong>png:</strong> 2"
],
"text/latex": [
"\\textbf{png:} 2"
],
"text/markdown": [
"**png:** 2"
],
"text/plain": [
"png \n",
" 2 "
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"list_clusters <- seq(5,15,5)\n",
"\n",
"# Save all boxplots in a single PDF output file\n",
"results_folder<-\"../results20210115154327/\"\n",
"# Save all boxplots in a single PDF output file\n",
"pdf(file=paste0(results_folder, \"simulated_boxplots.pdf\"), width = 15, height = 15, onefile = TRUE)\n",
"\n",
"# For each chosen number of clusters\n",
"for (i in list_clusters) {\n",
" \n",
" # Output files for each distribution\n",
" eq_file <- paste0(results_folder, i, \"_equal/\", \"JI.txt\")\n",
" het_file <- paste0(results_folder, i, \"_heterogeneous/\", \"JI.txt\")\n",
" if(exists(\"JI.final\")) rm(JI.final)\n",
" \n",
" # Load clusters (equal distribution)\n",
" if(file.exists(eq_file)) {\n",
" JI.final <- read.table(eq_file, sep=\"\\t\", header=TRUE)\n",
" names(JI.final) <- paste0(names(JI.final), \"_EQ\") \n",
" }\n",
" cat\n",
" # Load clusters (heterogeneous distribution)\n",
" if(file.exists(het_file)) {\n",
" JI.het <- read.table(het_file, sep=\"\\t\", header=TRUE)\n",
" names_methods <- names(JI.het)\n",
" names(JI.het) <- paste0(names(JI.het), \"_HET\")\n",
"\n",
" # Aggregate results\n",
" if(exists(\"JI.final\")) {\n",
" JI.final <- data.frame(JI.het, JI.final)\n",
" new_order <- apply(expand.grid(c(\"_HET\", \"_EQ\"), names_methods)[, c(2,1)], 1, paste, collapse=\"\")\n",
" JI.final <- JI.final[, new_order]\n",
" }\n",
" else {\n",
" JI.final <- JI.het\n",
" }\n",
" }\n",
" \n",
" # Plot results\n",
" par(mar=c(25,4,2,2)+.1)\n",
" boxplot(JI.final, xaxt=\"none\", cex.axis=3.5, \n",
" col=c('gray','gray','red','red','blue','blue','blueviolet','blueviolet','deeppink','deeppink','chocolate1','chocolate1','darkgoldenrod','darkgoldenrod','green','green','darkturquoise','darkturquoise'), \n",
" ann=FALSE, outline=FALSE)\n",
" matplot(1:ncol(JI.final), t(JI.final), col=\"black\", pch=16, xaxt=\"none\", cex=0.8, add=TRUE)\n",
" axis(1, at=1:ncol(JI.final), labels=names(JI.final), \n",
" las=2, srt=45, cex=0.8, cex.lab=3.5, cex.axis=3.5, cex.main=1.5, cex.sub=1.5) \n",
" title(main=paste(i,\"clusters\",sep=\" \"), \n",
" cex.lab=0.75, line = -2.5, adj=0, cex.main=3.5)\n",
"}\n",
"dev.off()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"list_clusters <- seq(5,15,5)\n",
"list_distrib <- c(\"heterogeneous\",\"equal\")\n",
"# Save all boxplots in a single PDF output file\n",
"results_folder<-\"../data/simulations_20210329212141/\"\n",
"pdf(file=paste0(results_folder, \"AE_simulated_boxplots2.pdf\"), width = 15, height = 15, onefile = TRUE)\n",
"\n",
"# For each chosen number of clusters\n",
"for (i in list_clusters) {\n",
" \n",
" # Output files for each distribution\n",
" eq_file <- paste0(results_folder, \"equal/\", i,\"/JI2.txt\")\n",
" het_file <- paste0(results_folder, \"heterogeneous/\",i, \"/JI2.txt\")\n",
" print(eq_file)\n",
" if(exists(\"JI.final\")) rm(JI.final)\n",
" \n",
" # Load clusters (equal distribution)\n",
" if(file.exists(eq_file)) {\n",
" JI.final <- read.table(eq_file, sep=\"\\t\", header=TRUE)\n",
" print(JI.final)\n",
" names(JI.final) <- paste0(names(JI.final), \"_EQ\") \n",
" }\n",
" #Load clusters (heterogeneous distribution)\n",
" if(file.exists(het_file)) {\n",
" JI.het <- read.table(het_file, sep=\"\\t\", header=TRUE)\n",
" names_methods <- names(JI.het)\n",
" names(JI.het) <- paste0(names(JI.het), \"_HET\")\n",
"\n",
" # Aggregate results\n",
" if(exists(\"JI.final\")) {\n",
" JI.final <- data.frame(JI.het, JI.final)\n",
" new_order <- apply(expand.grid(c(\"_HET\", \"_EQ\"), names_methods)[, c(2,1)], 1, paste, collapse=\"\")\n",
" JI.final <- JI.final[, new_order]\n",
" }\n",
" else {\n",
" JI.final <- JI.het\n",
" }\n",
" }\n",
" print(names(JI.final))\n",
" \n",
" labels<-list(\"AE_FAETC_HET\",\"AE_FAETC_EQ\",\"AE_FCTAE_HET\",\"AE_FCTAE_EQ\",\"DAE_FAETC_HET\",\"DAE_FAETC_EQ\",\"DAE_FCTAE_HET\",\"DAE_FCTAE_EQ\",\"VAE_FCTAE_HET\" \n",
",\"VAE_FCTAE_EQ\",\"SVAE_FCTAE_HET\",\"SVAE_FCTAE_EQ\",\"MMDVAE_HET\",\"MMDVAE_EQ\" )\n",
" # Plot results\n",
" par(mar=c(25,4,2,2)+.1)\n",
" boxplot(JI.final, xaxt=\"none\", cex.axis=3.5, \n",
" col=c('gray','gray','red','red','blue','blue','blueviolet','blueviolet','deeppink','deeppink','chocolate1','chocolate1'), \n",
" ann=FALSE, outline=FALSE)\n",
" matplot(1:ncol(JI.final), t(JI.final), col=\"black\", pch=16, xaxt=\"none\", cex=0.8, add=TRUE)\n",
" axis(1, at=1:ncol(JI.final), labels=labels, \n",
" las=2, srt=45, cex=0.5, cex.lab=2, cex.axis=2, cex.main=1.5, cex.sub=1.5) \n",
" title(main=paste(i,\"clusters\",sep=\" \"), \n",
" cex.lab=0.75, line = -2.5, adj=0, cex.main=3.5)\n",
"}\n",
"dev.off()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "R",
"language": "R",
"name": "ir"
},
"language_info": {
"codemirror_mode": "r",
"file_extension": ".r",
"mimetype": "text/x-r-source",
"name": "R",
"pygments_lexer": "r",
"version": "3.5.1"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Datasets
Models
