In jeksterslab/ramR: Reticular Action Model (RAM) Notation

knitr::opts_chunk$set(
  error = TRUE,
  collapse = TRUE,
  comment = "#>",
  out.width = "100%"
)

alpha <- runif(n = 1, min = -1, max = 1)
beta <- runif(n = 1, min = -1, max = 1)
sigmax2 <- runif(n = 1, min = 0, max = 1)
sigmae2 <- runif(n = 1, min = 0, max = 1)
mux <- runif(n = 1, min = -1, max = 1)
muepsilon <- 0
A <- S <- matrixR::ZeroMatrix(3)
A[1, ] <- c(0, beta, 1)
diag(S) <- c(0, sigmax2, sigmae2)
colnames(A) <- rownames(A) <- c("y", "x", "e")
I <- matrixR::IdentityFrom(A)
IminusA <- I - A
E <- Ryacas::as_r(solve(Ryacas::ysym(I) - Ryacas::ysym(A)))
Filter <- diag(2)
Filter <- cbind(Filter, 0)
colnames(Filter) <- c("y", "x", "e")
rownames(Filter) <- c("y", "x")
u <- as.matrix(c(alpha, mux, 0))
v <- as.matrix(c(alpha + (beta * mux), mux, 0.00))
rownames(u) <- rownames(v) <- c("y", "x", "e")
colnames(u) <- "u"
colnames(v) <- "v"
C <- Ryacas::as_r(
  solve(
    Ryacas::ysym(I) - Ryacas::ysym(A)
  ) * Ryacas::ysym(S) * t(
    solve(Ryacas::ysym(I) - Ryacas::ysym(A))
  )
)
C.scaled <- stats::cov2cor(C)
M <- Ryacas::as_r(
  Ryacas::ysym(Filter) * solve(
    Ryacas::ysym(I) - Ryacas::ysym(A)
  ) * Ryacas::ysym(S) * t(
    solve(Ryacas::ysym(I) - Ryacas::ysym(A))
  ) * t(Ryacas::ysym(Filter))
)
M.scaled <- Filter %*% C.scaled %*% t(Filter)
g <- as.matrix(
  Ryacas::as_r(
    Ryacas::ysym(Filter) * Ryacas::ysym(v)
  )
)

\begin{equation} \begin{split} y &= \alpha + \beta x + \varepsilon \ y &= r alpha + \left( r beta x \right) + \varepsilon \end{split} \end{equation}

n <- 100
Data <- ramR::RAM2Data(n, A, S, u, Filter, empirical = TRUE)
mu <- as.matrix(colMeans(Data))
Sigma <- cov(Data)
R <- cor(Data)

testthat::test_that("g.", {
  for (i in seq_len(nrow(g))) {
    for (j in seq_len(ncol(g))) {
      testthat::expect_equal(
        g[i, j],
        mu[i, j],
        check.attributes = FALSE
      )
    }
  }
})
testthat::test_that("M.", {
  for (i in seq_len(nrow(M))) {
    for (j in seq_len(ncol(M))) {
      testthat::expect_equal(
        M[i, j],
        Sigma[i, j],
        check.attributes = FALSE
      )
    }
  }
})
testthat::test_that("M.scaled", {
  for (i in seq_len(nrow(M.scaled))) {
    for (j in seq_len(ncol(M.scaled))) {
      testthat::expect_equal(
        M.scaled[i, j],
        R[i, j],
        check.attributes = FALSE
      )
    }
  }
})
testthat::test_that("n.", {
  testthat::expect_equal(
    n,
    dim(Data)[1],
    check.attributes = FALSE
  )
})

jeksterslab/ramR documentation built on March 14, 2021, 9:38 a.m.