Mx = 6 #мат ожидание признака
My = 13 #мат ожидание прогноз параметра
SigmX = 1 #дисперсия
SigmY = 0.5
R <- c(0.5, -0.9)
X <- c(16, 15, 14, 13)
DispX = SigmX^2
DispY = SigmY^2
#одномерная плотность распределения признака
W.x <- (1/(SigmX * sqrt(2*pi)) * exp(1)^(-((X-Mx)^2)/2*DispX))

#задание 3.3
Dy <- c(DispY, 2 * DispY, 4 * DispY)
#модуль R
r <- seq(from = 0, to = 1, by = 0.1)
#вектор дисперсий что ли
Dymtx <- matrix(Dy, nrow = 11, ncol = 3, byrow = TRUE)
for (i in 1:11) {
  Dymtx[i,] = Dymtx[i,] * (1 - r[i]^2)
}
Svodka <- data.frame("Dy" = Dymtx[,1], "Dy2" = Dymtx[,2], "Dy4" = Dymtx[,3])
View(Svodka)
plot(r, Dymtx[,1], type = "l",
     main = "Зависимость дипсперсии погрешности прогнозирования от коэффициента кореляции",
     xlab = "R", ylab = "Dy", ylim = c(0, 1))
text(r[3], 0.3, "Dy1")
lines(r, Dymtx[,2], type = "l")
text(r[5], 0.57, "Dy2")
lines(r, Dymtx[,3], type = "l")
text(r[7], 0.87, "Dy3")

#задание 3.4
coeffic <- c(-3:-2,seq(-1,1, by = 0.2),2:3)
W.y <- (1/(SigmY * sqrt(2*pi)) * exp(1)^(-(((My - coeffic*SigmY)-My)^2)/2*DispX))

#задание 3.5
My_xR1 <- (My + R[1] * (SigmY/SigmX)*(X-Mx))
My_xR2 <- (My + R[2] * (SigmY/SigmX)*(X-Mx))
Dy_xR1 = DispX * (1-R[1]^2)
Dy_xR2 = DispX * (1-R[2]^2)
Wy_xR1 <- matrix(NA, nrow = length(X), ncol = length(coeffic))
Wy_xR2 <- matrix(NA, nrow = length(X), ncol = length(coeffic))
for (i in X) {
  for (j in coeffic) {
      Wy_xR1[i][j] = 1/(SigmY * sqrt(1 - R[1]^2)* sqrt(2 * pi)) *
        exp(1)^(-((((My_xR1[i] - coeffic[j] * SigmY) - My_xR1[i])^2)/(2 * Dy_xR1)))
      Wy_xR2[i][j] = 1/(SigmY * sqrt(1 - R[2]^2)* sqrt(2 * pi)) *
        exp(1)^(-((((My_xR2[i] - coeffic[j] * SigmY) - My_xR2[i])^2)/(2 * Dy_xR2[i])))
  }
}