X-Git-Url: https://code.communitydata.science/ml_measurement_error_public.git/blobdiff_plain/588bdd7ed74cf8fe8fd0f15df58a6a40c26ebae5..2cd447c327744263d5f94b20e1146cdf31b2ec2c:/simulations/03_depvar_differential.R?ds=inline

diff --git a/simulations/03_depvar_differential.R b/simulations/03_depvar_differential.R
index d52afe7..872931f 100644
--- a/simulations/03_depvar_differential.R
+++ b/simulations/03_depvar_differential.R
@@ -31,18 +31,18 @@ source("simulation_base.R")
 
 ## one way to do it is by adding correlation to x.obs and y that isn't in w.
 ## in other words, the model is missing an important feature of x.obs that's related to y.
-simulate_data <- function(N, m, B0, Bxy, Bgy, seed, prediction_accuracy=0.73, accuracy_imbalance_difference=0.3){
+simulate_data <- function(N, m, B0, Bxy, Bzy, seed, prediction_accuracy=0.73, accuracy_imbalance_difference=0.3){
     set.seed(seed)
     # make w and y dependent
-    g <- rbinom(N, 1, 0.5)
+    z <- rbinom(N, 1, 0.5)
     x <- rbinom(N, 1, 0.5)
 
-    ystar <- Bgy * g + Bxy * x
-    y <- rbinom(N,1,logistic(ystar))
+    ystar <- Bzy * z + Bxy * x
+    y <- rbinom(N,1,plogis(ystar))
 
-    # glm(y ~ x + g, family="binomial")
+    # glm(y ~ x + z, family="binomial")
 
-    df <- data.table(x=x,y=y,ystar=ystar,g=g)
+    df <- data.table(x=x,y=y,ystar=ystar,z=z)
 
     if(m < N){
         df <- df[sample(nrow(df), m), y.obs := y]
@@ -52,36 +52,44 @@ simulate_data <- function(N, m, B0, Bxy, Bgy, seed, prediction_accuracy=0.73, ac
 
     df <- df[,w_pred:=y]
 
-    pg <- mean(g)
+    pz <- mean(z)
 
     accuracy_imbalance_ratio <- (prediction_accuracy + accuracy_imbalance_difference/2) / (prediction_accuracy - accuracy_imbalance_difference/2)
 
     # this works because of conditional probability
-    accuracy_g0 <- prediction_accuracy / (pg*(accuracy_imbalance_ratio) + (1-pg))
-    accuracy_g1 <- accuracy_imbalance_ratio * accuracy_g0
+    accuracy_z0 <- prediction_accuracy / (pz*(accuracy_imbalance_ratio) + (1-pz))
+    accuracy_z1 <- accuracy_imbalance_ratio * accuracy_z0
 
-    dfg0 <- df[g==0]
-    ng0 <- nrow(dfg0)
-    dfg1 <- df[g==1]
-    ng1 <- nrow(dfg1)
 
-    dfg0 <- dfg0[sample(ng0, (1-accuracy_g0)*ng0), w_pred := (w_pred-1)**2]
-    dfg1 <- dfg1[sample(ng1, (1-accuracy_g1)*ng1), w_pred := (w_pred-1)**2]
+    yz0 <- df[z==0]$y
+    yz1 <- df[z==1]$y
+    nz1 <- nrow(df[z==1])
+    nz0 <- nrow(df[z==0])
 
-    df <- rbind(dfg0,dfg1)
+    acc_z0 <- plogis(0.7*scale(yz0) + qlogis(accuracy_z0))
+    acc_z1 <- plogis(1.3*scale(yz1) + qlogis(accuracy_z1))
+    
+    w0z0 <- (1-yz0)**2 + (-1)**(1-yz0) * acc_z0
+    w0z1 <- (1-yz1)**2 + (-1)**(1-yz1) * acc_z1
+    
+    w0z0.noisy.odds <- rlogis(nz0,qlogis(w0z0))
+    w0z1.noisy.odds <- rlogis(nz1,qlogis(w0z1))
+    df[z==0,w:=plogis(w0z0.noisy.odds)]
+    df[z==1,w:=plogis(w0z1.noisy.odds)]
 
-    wmod <- glm(y.obs ~ w_pred,data=df[!is.null(y.obs)],family=binomial(link='logit'))
-    w <- predict(wmod,df,type='response')
+    df[,w_pred:=as.integer(w > 0.5)]
 
-    df <- df[,':='(w=w)]
+    print(mean(df[y==0]$y == df[y==0]$w_pred))
+    print(mean(df[y==1]$y == df[y==1]$w_pred))
+    print(mean(df$w_pred == df$y))
 
     return(df)
 }
 
 parser <- arg_parser("Simulate data and fit corrected models")
-parser <- add_argument(parser, "--N", default=5000, help="number of observations of w")
-parser <- add_argument(parser, "--m", default=200, help="m the number of ground truth observations")
-parser <- add_argument(parser, "--seed", default=4321, help='seed for the rng')
+parser <- add_argument(parser, "--N", default=1000, help="number of observations of w")
+parser <- add_argument(parser, "--m", default=500, help="m the number of ground truth observations")
+parser <- add_argument(parser, "--seed", default=17, help='seed for the rng')
 parser <- add_argument(parser, "--outfile", help='output file', default='example_2.feather')
 parser <- add_argument(parser, "--y_explained_variance", help='what proportion of the variance of y can be explained?', default=0.005)
 parser <- add_argument(parser, "--prediction_accuracy", help='how accurate is the predictive model?', default=0.73)
@@ -90,24 +98,26 @@ parser <- add_argument(parser, "--accuracy_imbalance_difference", help='how much
 args <- parse_args(parser)
 
 B0 <- 0
-Bxy <- 0.2
-Bgy <- -0.2
+Bxy <- 0.7
+Bzy <- -0.7
 
-df <- simulate_data(args$N, args$m, B0, Bxy, Bgy, args$seed, args$prediction_accuracy, args$accuracy_imbalance_difference)
+if(args$m < args$N){
+    df <- simulate_data(args$N, args$m, B0, Bxy, Bzy, args$seed, args$prediction_accuracy, args$accuracy_imbalance_difference)
 
-result <- list('N'=args$N,'m'=args$m,'B0'=B0,'Bxy'=Bxy,'Bgy'=Bgy, 'seed'=args$seed, 'y_explained_variance'=args$y_explained_variance, 'prediction_accuracy'=args$prediction_accuracy, 'accuracy_imbalance_difference'=args$accuracy_imbalance_difference)
+    result <- list('N'=args$N,'m'=args$m,'B0'=B0,'Bxy'=Bxy,'Bzy'=Bzy, 'seed'=args$seed, 'y_explained_variance'=args$y_explained_variance, 'prediction_accuracy'=args$prediction_accuracy, 'accuracy_imbalance_difference'=args$accuracy_imbalance_difference)
 
-outline <- run_simulation_depvar(df=df, result)
+    outline <- run_simulation_depvar(df, result, outcome_formula = y ~ x + z, proxy_formula = w_pred ~ y*x + y*z + z*x)
 
+    outfile_lock <- lock(paste0(args$outfile, '_lock'),exclusive=TRUE)
 
-outfile_lock <- lock(paste0(args$outfile, '_lock'),exclusive=TRUE)
-if(file.exists(args$outfile)){
-    logdata <- read_feather(args$outfile)
-    logdata <- rbind(logdata,as.data.table(outline))
-} else {
-    logdata <- as.data.table(outline)
-}
+    if(file.exists(args$outfile)){
+        logdata <- read_feather(args$outfile)
+        logdata <- rbind(logdata,as.data.table(outline),fill=TRUE)
+    } else {
+        logdata <- as.data.table(outline)
+    }
 
-print(outline)
-write_feather(logdata, args$outfile)
-unlock(outfile_lock)
+    print(outline)
+    write_feather(logdata, args$outfile)
+    unlock(outfile_lock)
+}