# pi00: Estimates of pi_00, the Proportion of Simultaneous True Nulls... In ntyet/sfdr: False Discovery Rate Estimation in Two Independent Experiments

## Description

This function estimate Proportion of Simultaneous True Nulls in Two Independent Experiments from the inputs including 2 vectors of p-values.

## Usage

 `1` ```pi00(p1, p2, lambda1, lambda2) ```

## Arguments

 `p1, p2` vectors of p-values of the two independent experiments. `lambda1, lambda2` lambda thresholds for the two experiments.

## Value

The function returns an estimate of the proportion of the number of true simultaneous nulls. The value is a real number between 0 and 1.

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14``` ```n <- 1000 n11 <- 25; n10 <- n01 <- 0; n00 <- n - (n11 + n10 + n01) mu1 <- rep(0, n) mu1[1:(n11+n10)] <- rnorm(n = n11+n10, mean = 6,sd = 1) t1 <- plyr::laply(1:n, function(i)rt(n = 1, df = 4, ncp = mu1[i])) p1 <- plyr::laply(1:n, function(i)2*pt(q = abs(t1[i]), df = 4, ncp = 0, lower.tail = FALSE)) lambda1 <- estimate.lambda(p1, B = 20) mu2 <- rep(0, n) mu2[c(1:n11, (n11+n10+1):(n11+n10+n01))] <- rnorm(n = n11+n01, mean = 6,sd = 1) t2 <- plyr::laply(1:n, function(i)rt(n = 1, df = 4, ncp = mu2[i])) p2 <- plyr::laply(1:n, function(i)2*pt(q = abs(t2[i]), df = 4, ncp = 0, lower.tail = FALSE)) lambda2 <- estimate.lambda(p2, B = 20) pi_00 <- pi00(p1, p2, lambda1, lambda2) pi_00 ```

ntyet/sfdr documentation built on Jan. 13, 2018, 6:50 a.m.