New covariance function in rstan

Summary:

Based on my understanding of https://github.com/stan-dev/stan/pull/2739, there should be new covariance function available to use via rstan.

Description:

It seems that these functions are not yet available.

Reproducible Steps:

Simple GP simulation model from the Stan manual:

data {

	int<lower=1> N;

	int<lower=1> D; //dimension of the GP
	vector[D] x[N]; //point where the GP is estimated
}

transformed data {
	matrix[N, N] L; //cholesky matrix
        vector[N] mu = rep_vector(0, N); //assume 0-mean
	matrix[N, N] K = gp_matern32_cov(x, 1.0, 1.0);

	for(n in 1:N)

		K[n, n] = K[n, n] + 0.1; // to ensure positive-definite matrix
	L = cholesky_decompose(K);

}

parameters {

	vector[N] eta;

}

model {

	eta ~ std_normal();

}
generated quantities {
	vector[N] y;
	y = mu + L * eta;
}

And in R:

N <- 100
D <- 2
x <- matrix(runif(N * D), nrow = N, ncol = D)
m <- stan("GP_mat32_sim.stan", data = list(x = x, N = 100, D = 2))

Current Output:

SYNTAX ERROR, MESSAGE(S) FROM PARSER:
No matches for:

gp_matern32_cov(vector[ ], real, real)

Function gp_matern32_cov not found.
error in 'model1b50107f4b65_GP_mat32_sim' at line 8, column 49

 6: transformed data {
 7: 	matrix[N, N] L; //cholesky matrix
 8: 	matrix[N, N] K = gp_matern32_cov(x, 1.0, 1.0);
                                                    ^
 9: 	for(n in 1:N)

---

Error in stanc(file = file, model_code = model_code, model_name = model_name, :
failed to parse Stan model 'GP_mat32_sim' due to the above error.

Expected Output:

NA

RStan Version:

rstan_2.19.2

R Version:

R version 3.6.1 (2019-07-05)

Operating System:

Win10

Start with the GP_mat32_sim.stan example and the rstan_2.19.2 parser path, then compare the covariance functions described in Stan pull request 2739 with those available through rstan. Reproduce the gp_matern32_cov parsing error using stanc and confirm completion when the example model parses and runs through the R stan() call.