Fix typos in comments and variable names (#2665)

* Initial plan

* Fix typos in comments and code

Co-authored-by: yebai <[email protected]>

---------

Co-authored-by: copilot-swe-agent[bot] <[email protected]>
Co-authored-by: yebai <[email protected]>

Author: Copilot

src/mcmc/Inference.jl

@@ -81,7 +81,7 @@ export InferenceAlgorithm,
     externalsampler
 
 ###############################################
-# Abstract inferface for inference algorithms #
+# Abstract interface for inference algorithms #
 ###############################################
 
 include("algorithm.jl")

src/mcmc/gibbs.jl

@@ -601,7 +601,7 @@ end
     match_linking!!(varinfo_local, prev_state_local, model)
 
 Make sure the linked/invlinked status of varinfo_local matches that of the previous
-state for this sampler. This is relevant when multilple samplers are sampling the same
+state for this sampler. This is relevant when multiple samplers are sampling the same
 variables, and one might need it to be linked while the other doesn't.
 """
 function match_linking!!(varinfo_local, prev_state_local, model)

src/mcmc/particle_mcmc.jl

@@ -333,8 +333,8 @@ function DynamicPPL.initialstep(
 
     # Pick a particle to be retained.
     Ws = AdvancedPS.getweights(particles)
-    indx = AdvancedPS.randcat(rng, Ws)
-    reference = particles.vals[indx]
+    index = AdvancedPS.randcat(rng, Ws)
+    reference = particles.vals[index]
 
     # Compute the first transition.
     _vi = reference.model.f.varinfo
@@ -373,8 +373,8 @@ function AbstractMCMC.step(
 
     # Pick a particle to be retained.
     Ws = AdvancedPS.getweights(particles)
-    indx = AdvancedPS.randcat(rng, Ws)
-    newreference = particles.vals[indx]
+    index = AdvancedPS.randcat(rng, Ws)
+    newreference = particles.vals[index]
 
     # Compute the transition.
     _vi = newreference.model.f.varinfo

test/ext/OptimInterface.jl

@@ -134,7 +134,7 @@ using Turing
     # Some of the models have one variance parameter per observation, and so
     # the MLE should have the variances set to 0. Since we're working in
     # transformed space, this corresponds to `-Inf`, which is of course not achievable.
-    # In particular, it can result in "early termniation" of the optimization process
+    # In particular, it can result in "early termination" of the optimization process
     # because we hit NaNs, etc. To avoid this, we set the `g_tol` and the `f_tol` to
     # something larger than the default.
     allowed_incorrect_mle = [

test/mcmc/gibbs.jl

@@ -383,7 +383,7 @@ end
         # As above but different samplers and using kwargs.
         s3 = Gibbs(:s => CSMC(3), :m => HMCDA(200, 0.65, 0.15))
         s4 = Gibbs(@varname(s) => HMC(0.1, 5), @varname(m) => ESS())
-        # Multiple instnaces of the same sampler. This implements running, in this case,
+        # Multiple instances of the same sampler. This implements running, in this case,
         # 3 steps of HMC on m and 2 steps of PG on m in every iteration of Gibbs.
         s5 = begin
             hmc = HMC(0.1, 5)

test/optimisation/Optimisation.jl

@@ -505,7 +505,7 @@ using Turing
     # Some of the models have one variance parameter per observation, and so
     # the MLE should have the variances set to 0. Since we're working in
     # transformed space, this corresponds to `-Inf`, which is of course not achievable.
-    # In particular, it can result in "early termniation" of the optimization process
+    # In particular, it can result in "early termination" of the optimization process
     # because we hit NaNs, etc. To avoid this, we set the `g_tol` and the `f_tol` to
     # something larger than the default.
     allowed_incorrect_mle = [