Uninformative error message with incorrect usage of 'fit' argument

Summary:

Passing a string or model code to the fit argument results in an error about the wrong algorithm being used, rather than the wrong argument.

Description:

As discovered by a user on the forum.

When calling: stan(fit = stan_code,...) where stan_code is a string (rather than stanfit object), the model compiles and then returns an error about needing to use the fixed_param algorithm. This doesn't make it clear the problem is with the argument to stan() and not the choice of sampler.

Reproducible Steps:

library(rstan) 
stancode <- 'data {real y_mean;} parameters {real y;} model {y ~ normal(y_mean,1);}'
out = stan(fit = stancode,data=list(y_mean=0))

SAMPLING FOR MODEL '81051bcc2cf1bedf378224b0a93e2877' NOW (CHAIN 1).
[1] "Error in sampler$call_sampler(args_list[[i]]) : "                      
[2] "  Must use algorithm=\"Fixed_param\" for model that has no parameters."
[1] "error occurred during calling the sampler; sampling not done"

RStan Version:

2.19.2

R Version:

R version 3.6.1 (2019-07-05)

Operating System:

Windows 10

Start at the stan() entry point and reproduce the reported call with a string passed to fit, using the RStan and R versions shown. Trace where the fit argument is handled before sampling. Done means the same misuse reports that fit must be a stanfit object, rather than suggesting the fixed_param algorithm.