simulations/plot_example.R

   1 source("RemembR/R/RemembeR.R")
   2 library(arrow)
   3 library(data.table)
   4 library(ggplot2)
   5 library(filelock)
   6 library(argparser)
   7
   8 parser <- arg_parser("Simulate data and fit corrected models.")
   9 parser <- add_argument(parser, "--infile", default="", help="name of the file to read.")
  10 parser <- add_argument(parser, "--name", default="", help="The name to safe the data to in the remember file.")
  11 args <- parse_args(parser)
  12
  13 build_plot_dataset <- function(df){
  14     x.naive <- df[,.(N, m, Bxy, Bxy.est.naive, Bxy.ci.lower.naive, Bxy.ci.upper.naive)]
  15     x.naive <- x.naive[,':='(true.in.ci = as.integer((Bxy >= Bxy.ci.lower.naive) & (Bxy <= Bxy.ci.upper.naive)),
  16                              zero.in.ci = (0 >= Bxy.ci.lower.naive) & (0 <= Bxy.ci.upper.naive),
  17                              bias = Bxy - Bxy.est.naive,
  18                              Bxy.est.naive = Bxy.est.naive,
  19                              sign.correct = as.integer(sign(Bxy) == sign(Bxy.est.naive)))]
  20
  21     x.naive.plot <- x.naive[,.(p.true.in.ci = mean(true.in.ci),
  22                                mean.bias = mean(bias),
  23                                mean.est = mean(Bxy.est.naive),
  24                                var.est = var(Bxy.est.naive),
  25                                N.sims = .N,
  26                                p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
  27                                variable='x',
  28                                method='Naive'
  29                                ),
  30                             by=c('N','m')]
  31
  32
  33     g.naive <- df[,.(N, m, Bgy, Bgy.est.naive, Bgy.ci.lower.naive, Bgy.ci.upper.naive)]
  34     g.naive <- g.naive[,':='(true.in.ci = as.integer((Bgy >= Bgy.ci.lower.naive) & (Bgy <= Bgy.ci.upper.naive)),
  35                              zero.in.ci = (0 >= Bgy.ci.lower.naive) & (0 <= Bgy.ci.upper.naive),
  36                              bias = Bgy - Bgy.est.naive,
  37                              Bgy.est.naive = Bgy.est.naive,
  38                              sign.correct = as.integer(sign(Bgy) == sign(Bgy.est.naive)))]
  39
  40     g.naive.plot <- g.naive[,.(p.true.in.ci = mean(true.in.ci),
  41                                mean.bias = mean(bias),
  42                                mean.est = mean(Bgy.est.naive),
  43                                var.est = var(Bgy.est.naive),
  44                                N.sims = .N,
  45                                p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
  46                                variable='g',
  47                                method='Naive'
  48                                ),
  49                             by=c('N','m')]
  50
  51
  52     x.feasible <- df[,.(N, m, Bxy, Bxy.est.feasible, Bxy.ci.lower.feasible, Bxy.ci.upper.feasible)]
  53     x.feasible <- x.feasible[,':='(true.in.ci = as.integer((Bxy >= Bxy.ci.lower.feasible) & (Bxy <= Bxy.ci.upper.feasible)),
  54                                    zero.in.ci = (0 >= Bxy.ci.lower.feasible) & (0 <= Bxy.ci.upper.feasible),
  55                                    bias = Bxy - Bxy.est.feasible,
  56                                    Bxy.est.feasible = Bxy.est.feasible,
  57                                    sign.correct = as.integer(sign(Bxy) == sign(Bxy.est.feasible)))]
  58
  59     x.feasible.plot <- x.feasible[,.(p.true.in.ci = mean(true.in.ci),
  60                                      mean.bias = mean(bias),
  61                                      mean.est = mean(Bxy.est.feasible),
  62                                      var.est = var(Bxy.est.feasible),
  63                                      N.sims = .N,
  64                                      p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
  65                                      variable='x',
  66                                      method='Feasible'
  67                                      ),
  68                                   by=c('N','m')]
  69
  70
  71     g.feasible <- df[,.(N, m, Bgy, Bgy.est.feasible, Bgy.ci.lower.feasible, Bgy.ci.upper.feasible)]
  72     g.feasible <- g.feasible[,':='(true.in.ci = as.integer((Bgy >= Bgy.ci.lower.feasible) & (Bgy <= Bgy.ci.upper.feasible)),
  73                                    zero.in.ci = (0 >= Bgy.ci.lower.feasible) & (0 <= Bgy.ci.upper.feasible),
  74                                    bias = Bgy - Bgy.est.feasible,
  75                                    Bgy.est.feasible = Bgy.est.feasible,
  76                                    sign.correct = as.integer(sign(Bgy) == sign(Bgy.est.feasible)))]
  77
  78     g.feasible.plot <- g.feasible[,.(p.true.in.ci = mean(true.in.ci),
  79                                      mean.bias = mean(bias),
  80                                      mean.est = mean(Bgy.est.feasible),
  81                                      var.est = var(Bgy.est.feasible),
  82                                      N.sims = .N,
  83                                      p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
  84                                      variable='g',
  85                                      method='Feasible'
  86                                      ),
  87                                   by=c('N','m')]
  88
  89
  90
  91     x.amelia.full <- df[,.(N, m, Bxy, Bxy.est.true, Bxy.ci.lower.amelia.full, Bxy.ci.upper.amelia.full, Bxy.est.amelia.full)]
  92
  93     x.amelia.full <- x.amelia.full[,':='(true.in.ci = (Bxy.est.true >= Bxy.ci.lower.amelia.full) & (Bxy.est.true <= Bxy.ci.upper.amelia.full),
  94                                          zero.in.ci = (0 >= Bxy.ci.lower.amelia.full) & (0 <= Bxy.ci.upper.amelia.full),
  95                                          bias = Bxy.est.true - Bxy.est.amelia.full,
  96                                          sign.correct = sign(Bxy.est.true) == sign(Bxy.est.amelia.full))]
  97
  98     x.amelia.full.plot <- x.amelia.full[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
  99                                            mean.bias = mean(bias),
 100                                            p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 101                                            mean.est = mean(Bxy.est.amelia.full),
 102                                            var.est = var(Bxy.est.amelia.full),
 103                                            N.sims = .N,
 104                                            variable='x',
 105                                            method='Multiple imputation'
 106                                            ),
 107                                         by=c('N','m')]
 108
 109
 110     g.amelia.full <- df[,.(N, m, Bgy.est.true, Bgy.est.amelia.full, Bgy.ci.lower.amelia.full, Bgy.ci.upper.amelia.full)]
 111     g.amelia.full <- g.amelia.full[,':='(true.in.ci = (Bgy.est.true >= Bgy.ci.lower.amelia.full) & (Bgy.est.true <= Bgy.ci.upper.amelia.full),
 112                                          zero.in.ci = (0 >= Bgy.ci.lower.amelia.full) & (0 <= Bgy.ci.upper.amelia.full),
 113                                          bias =  Bgy.est.amelia.full - Bgy.est.true,
 114                                          sign.correct = sign(Bgy.est.true) == sign(Bgy.est.amelia.full))]
 115
 116     g.amelia.full.plot <- g.amelia.full[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 117                                            mean.bias = mean(bias),
 118                                            p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 119                                            mean.est = mean(Bgy.est.amelia.full),
 120                                            var.est = var(Bgy.est.amelia.full),
 121                                            N.sims = .N,
 122                                            variable='g',
 123                                            method='Multiple imputation'
 124                                            ),
 125                                         by=c('N','m')]
 126
 127     ## x.amelia.nok <- df[,.(N, m, Bxy.est.true, Bxy.est.amelia.nok, Bxy.ci.lower.amelia.nok, Bxy.ci.upper.amelia.nok)]
 128     ## x.amelia.nok <- x.amelia.nok[,':='(true.in.ci = (Bxy.est.true >= Bxy.ci.lower.amelia.nok) & (Bxy.est.true <= Bxy.ci.upper.amelia.nok),
 129     ##                                    zero.in.ci = (0 >= Bxy.ci.lower.amelia.nok) & (0 <= Bxy.ci.upper.amelia.nok),
 130     ##                                    bias =  Bxy.est.amelia.nok - Bxy.est.true,
 131     ##                                    sign.correct = sign(Bxy.est.true) == sign(Bxy.est.amelia.nok))]
 132
 133     ## x.amelia.nok.plot <- x.amelia.nok[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 134     ##                                      mean.bias = mean(bias),
 135     ##                                      p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 136     ##                                      mean.est = mean(Bxy.est.amelia.nok),
 137     ##                                      var.est = var(Bxy.est.amelia.nok),
 138     ##                                      N.sims = .N,
 139     ##                                      variable='x',
 140     ##                                      method='Multiple imputation (Classifier features unobserved)'
 141     ##                                      ),
 142     ##                                   by=c('N','m')]
 143
 144     ## g.amelia.nok <- df[,.(N, m, Bgy.est.true, Bgy.est.amelia.nok, Bgy.ci.lower.amelia.nok, Bgy.ci.upper.amelia.nok)]
 145     ## g.amelia.nok <- g.amelia.nok[,':='(true.in.ci = (Bgy.est.true >= Bgy.ci.lower.amelia.nok) & (Bgy.est.true <= Bgy.ci.upper.amelia.nok),
 146     ##                                    zero.in.ci = (0 >= Bgy.ci.lower.amelia.nok) & (0 <= Bgy.ci.upper.amelia.nok),
 147     ##                                    bias =  Bgy.est.amelia.nok - Bgy.est.true,
 148     ##                                    sign.correct = sign(Bgy.est.true) == sign(Bgy.est.amelia.nok))]
 149
 150     ## g.amelia.nok.plot <- g.amelia.nok[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 151     ##                                      mean.bias = mean(bias),
 152     ##                                      p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 153     ##                                      mean.est = mean(Bgy.est.amelia.nok),
 154     ##                                      var.est = var(Bgy.est.amelia.nok),
 155     ##                                      N.sims = .N,
 156     ##                                      variable='g',
 157     ##                                      method="Multiple imputation (Classifier features unobserved)"
 158     ##                                      ),
 159     ##                                   by=c('N','m')]
 160
 161
 162     x.mecor <- df[,.(N,m,Bxy.est.true, Bxy.est.mecor,Bxy.lower.mecor, Bxy.upper.mecor)]
 163
 164     x.mecor <- x.mecor[,':='(true.in.ci = (Bxy.est.true >= Bxy.lower.mecor) & (Bxy.est.true <= Bxy.upper.mecor),
 165                              zero.in.ci = (0 >= Bxy.lower.mecor) & (0 <= Bxy.upper.mecor),
 166                              bias =  Bxy.est.mecor - Bxy.est.true,
 167                              sign.correct = sign(Bxy.est.true) == sign(Bxy.est.mecor))]
 168
 169     x.mecor.plot <- x.mecor[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 170                                mean.bias = mean(bias),
 171                                p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 172                                mean.est = mean(Bxy.est.mecor),
 173                                var.est = var(Bxy.est.mecor),
 174                                N.sims = .N,
 175                                variable='x',
 176                                method='Regression Calibration'
 177                                ),
 178                             by=c('N','m')]
 179
 180     g.mecor <- df[,.(N,m,Bgy.est.true, Bgy.est.mecor,Bgy.lower.mecor, Bgy.upper.mecor)]
 181
 182     g.mecor <- g.mecor[,':='(true.in.ci = (Bgy.est.true >= Bgy.lower.mecor) & (Bgy.est.true <= Bgy.upper.mecor),
 183                              zero.in.ci = (0 >= Bgy.lower.mecor) & (0 <= Bgy.upper.mecor),
 184                              bias =  Bgy.est.mecor - Bgy.est.true,
 185                              sign.correct = sign(Bgy.est.true) == sign(Bgy.est.mecor))]
 186
 187     g.mecor.plot <- g.mecor[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 188                                mean.bias = mean(bias),
 189                                p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 190                                mean.est = mean(Bgy.est.mecor),
 191                                var.est = var(Bgy.est.mecor),
 192                                N.sims = .N,
 193                                variable='g',
 194                                method='Regression Calibration'
 195                                ),
 196                             by=c('N','m')]
 197
 198     ## x.mecor <- df[,.(N,m,Bgy.est.true, Bgy.est.mecor,Bgy.ci.lower.mecor, Bgy.ci.upper.mecor)]
 199
 200     ## x.mecor <- x.mecor[,':='(true.in.ci = (Bgy.est.true >= Bgy.ci.lower.mecor) & (Bgy.est.true <= Bgy.ci.upper.mecor),
 201     ##                                      zero.in.ci = (0 >= Bgy.ci.lower.mecor) & (0 <= Bgy.ci.upper.mecor),
 202     ##                                      bias =  Bgy.est.mecor - Bgy.est.true,
 203     ##                                      sign.correct = sign(Bgy.est.true) == sign(Bgy.est.mecor))]
 204
 205     ## x.mecor.plot <- x.mecor[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 206     ##                                        mean.bias = mean(bias),
 207     ##                                      p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 208     ##                                      variable='g',
 209     ##                                      method='Regression Calibration'
 210     ##                                      ),
 211     ##                                     by=c('N','m')]
 212
 213
 214     x.gmm <- df[,.(N,m,Bxy.est.true, Bxy.est.gmm,Bxy.ci.lower.gmm, Bxy.ci.upper.gmm)]
 215     x.gmm <- x.gmm[,':='(true.in.ci = (Bxy.est.true >= Bxy.ci.lower.gmm) & (Bxy.est.true <= Bxy.ci.upper.gmm),
 216                          zero.in.ci = (0 >= Bxy.ci.lower.gmm) & (0 <= Bxy.ci.upper.gmm),
 217                          bias =  Bxy.est.gmm - Bxy.est.true,
 218                          sign.correct = sign(Bxy.est.true) == sign(Bxy.est.gmm))]
 219
 220     x.gmm.plot <- x.gmm[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 221                            mean.bias = mean(bias),
 222                            p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 223                            mean.est = mean(Bxy.est.gmm),
 224
 225                            var.est = var(Bxy.est.gmm),
 226                            N.sims = .N,
 227                            variable='x',
 228                            method='2SLS+gmm'
 229                            ),
 230                         by=c('N','m')]
 231
 232     g.gmm <- df[,.(N,m,Bgy.est.true, Bgy.est.gmm,Bgy.ci.lower.gmm, Bgy.ci.upper.gmm)]
 233     g.gmm <- g.gmm[,':='(true.in.ci = (Bgy.est.true >= Bgy.ci.lower.gmm) & (Bgy.est.true <= Bgy.ci.upper.gmm),
 234                          zero.in.ci = (0 >= Bgy.ci.lower.gmm) & (0 <= Bgy.ci.upper.gmm),
 235                          bias =  Bgy.est.gmm - Bgy.est.true,
 236                          sign.correct = sign(Bgy.est.true) == sign(Bgy.est.gmm))]
 237
 238     g.gmm.plot <- g.gmm[,.(p.true.in.ci = mean(as.integer(true.in.ci)),
 239                            mean.bias = mean(bias),
 240                            p.sign.correct = mean(as.integer(sign.correct & (! zero.in.ci))),
 241                            mean.est = mean(Bgy.est.gmm),
 242                            var.est = var(Bgy.est.gmm),
 243                            N.sims = .N,
 244                            variable='g',
 245                            method='2SLS+gmm'
 246                            ),
 247                         by=c('N','m')]
 248
 249     accuracy <- df[,mean(accuracy)]
 250
 251     plot.df <- rbindlist(list(x.naive.plot,g.naive.plot,x.amelia.full.plot,g.amelia.full.plot,x.mecor.plot, g.mecor.plot, x.gmm.plot, g.gmm.plot, x.feasible.plot, g.feasible.plot),use.names=T)
 252
 253     plot.df[,accuracy := accuracy]
 254
 255     plot.df <- plot.df[,":="(sd.est=sqrt(var.est)/N.sims)]
 256
 257     return(plot.df)
 258 }
 259
 260
 261 df <- read_feather(args$infile)
 262 plot.df <- build_plot_dataset(df)
 263 remember(plot.df,args$name)
 264
 265
 266 ## df[gmm.ER_pval<0.05]
 267
 268
 269
 270
 271
 272 ## ggplot(plot.df[variable=='x'], aes(y=mean.est, ymax=mean.est + var.est/2, ymin=mean.est-var.est/2, x=method)) + geom_pointrange() + facet_grid(-m~N) + scale_x_discrete(labels=label_wrap_gen(10))
 273
 274 ## ggplot(plot.df,aes(y=N,x=m,color=p.sign.correct)) + geom_point() + facet_grid(variable ~ method) + scale_color_viridis_c(option='D') + theme_minimal() + xlab("Number of gold standard labels") + ylab("Total sample size")
 275
 276 ## ggplot(plot.df,aes(y=N,x=m,color=abs(mean.bias))) + geom_point() + facet_grid(variable ~ method) + scale_color_viridis_c(option='D') + theme_minimal() + xlab("Number of gold standard labels") + ylab("Total sample size")