stonegold546

Keybase proof

I hereby claim:

• I am stonegold546 on github.
• I am jamesuanhoro (https://keybase.io/jamesuanhoro) on keybase.
• I have a public key ASDQwPwPar3wn53oAgVAk6oGl6VZsS7Xlucm8AE_Rkgx3wo

To claim this, I am signing this object:

stonegold546

title Using our API: details at /api/v1/routes

note over API: LOGIN
note right of API: GET /api/v1/client_id
Dev->API: Request for Google OAuth Client ID
API->Dev: Returns Google OAuth Client ID as URL Dev can call
Dev->*Google: Request callback code from Google
Google-->Dev: Returns callback code
destroy Google
note right of API: GET /api/v1/use_callback_code

stonegold546

get_gci_fcn <- function(mymeansqsvec, MC = 100000, B, L, R, alpha = 0.05) {

  # Precondition:  mymeansqsvec is a vector of mean squares from
  #                the ANOVA fit of a two-way crossed models.  B>2
  #                is the number of levels of the first factor and
  #                L>2 is the number of levels of the second factor and
  #                R>0 is the number of replicates per cell.  A two-way
  #                balanced layout without interaction is assumed.
  #                MC is the number of Monte Carlo used to construct the
  #                GCI, with default one-hundred thousand.

stonegold546

# require(lme4)
mymod <- function(data) {
  result <- lme4::lmer(y ~ 1 + (1 | x), data = data, REML = FALSE)
  sumy <- summary(result)
  varb <- sumy$varcor$x[1]
  varw <- sumy$sigma ^ 2
  icc <- varb / (varb + varw)
  c(icc, varb, varw)
}
# mymod(data)

stonegold546

library(lavaan)
library(rstan)
library(rethinking)
library(loo)

options(mc.cores = parallel::detectCores())
rstan_options(auto_write = TRUE)

dat <- dplyr::select(HolzingerSwineford1939, x1:x9)

stonegold546

library(lavaan)
library(psych)

sim.fun <- function (lv, lambda, nrep = 2e3) {
  np <- nrow(lv)
  t(replicate(nrep, {
    X <- lv %*% lambda +
      matrix(rnorm(np * length(lambda), 0, sqrt(1 - lambda ^ 2)), np, byrow = TRUE)
    # summary(cfa(paste("F =~", paste0("V", 1:length(lambda), collapse = " + ")), X, std.lv = TRUE))
    

stonegold546

library(lavaan)

summary(m0 <- cfa(
  "F1 =~ x1 + x2 + x3 + x9\n F2 =~ x4 + x5 + x6\n F3 =~ x7 + x8 + x9\n x3 ~~ x5\n x2 ~~ x7\n x4 ~~ x7",
  HolzingerSwineford1939, std.lv = TRUE, meanstructure = TRUE))

library(rstan)

options(mc.cores = parallel::detectCores())
rstan_options(auto_write = TRUE)

stonegold546

data {
  real<lower = 0> sd_m;
  real<lower = 0> sd_m_diff;
  real<lower = 0> sd_st;
  real<lower = 0> sd_st_r;
  int<lower = 0, upper = 1> nu_choice;
  int<lower = 0> N;
  vector<lower = 0, upper = 1>[N] x;
  vector[N] y;
}

stonegold546

progesterone <- c(1513, 2025)
placebo <- c(1459, 2013)
(tab <- as.data.frame(rbind(progesterone, placebo)))
names(tab) <- c("pass", "total")
tab$fail <- tab$total - tab$pass
tab$treatment <- c(1, 0)
chisq.test(tab[c("pass", "fail")], correct = FALSE)
# chisq.test(long.tab$treatment, long.tab$pass, correct = FALSE)
summary(glm(cbind(pass, fail) ~ treatment, binomial, tab))

stonegold546

library(lavaan)

X <- HolzingerSwineford1939[, paste0("x", 1:9)]
colMeans(X)
apply(X, 2, sd)

# min-max scaling version
X.p <- as.data.frame(apply(X, 2, function (x) (x - min(x)) / (max(x) - min(x))))
colMeans(X.p)
apply(X.p, 2, sd)