Fix broken links

src/cmdstan-guide/external_code.Rmd

@@ -90,7 +90,7 @@ External C++ functions are currently the only way to encode a function with a
 known analytic gradient outside the Stan Math Library. This is done very
 similarly to how a function would be added to the Math library with a
 reverse-mode specialization. The following code is adapted from the [Stan Math
-documentation](https://mc-stan.org/math/getting_started.html).
+documentation](https://mc-stan.org/math/md_doxygen_2contributor__help__pages_2getting__started.html).
 
 Suppose you have the following (nonsensical) model which relies on a function
 called `my_dot_self`. We will implement this as a copy of the built-in

src/cmdstan-guide/log_prob_config.Rmd

@@ -20,8 +20,8 @@ This method takes 3 arguments:
 
 - `constrained_params` - Input file of parameters values on the constrained scale.
 A single set of constrained parameters can be specified using
-[JSON](json.html) or [Rdump](rdump.html) format.
-Alternatively, the input file can be set of draws in [StanCSV](stan_csv.html) format.
+[JSON](#json) or [Rdump](#rdump) format.
+Alternatively, the input file can be set of draws in [StanCSV](#stan_csv) format.
 
 
 - `unconstrained_params` - Input file (JSON or R dump) of parameter values

src/cmdstan-guide/optimize_config.Rmd

@@ -62,7 +62,7 @@ and also much faster than the other optimizers.
 but has the advantage of setting its own stepsize.
 
 See the Stan Reference Manual's
-[Optimization chapter](https://mc-stan.org/docs/reference-manual/optimization-algorithms.html) 
+[Optimization chapter](https://mc-stan.org/docs/reference-manual/optimization.html)
 for a description of these algorithms.
 
 All of the optimizers stream per-iteration intermediate approximations to the command line console.

src/cmdstan-guide/variational_config.Rmd

@@ -11,7 +11,7 @@ ADVI optimizes the ELBO in the real-coordinate space using
 [stochastic gradient ascent](https://mc-stan.org/docs/reference-manual/stochastic-gradient-ascent.html).
 The measures of convergence are similar to the
 relative tolerance scheme of Stan's
-[optimization algorithms](https://mc-stan.org/docs/reference-manual/optimization-algorithms.html).
+[optimization algorithms](https://mc-stan.org/docs/reference-manual/optimization.html).
 
 The algorithm progression consists of an adaptation phase followed by a sampling phase.
 The adaptation phase finds a good value for the step size scaling parameter `eta`.