remove old docs

docs/make.jl

@@ -29,15 +29,16 @@ makedocs(;
         "Home" => "index.md",
         "Walk Through" => "walk_through.md",
         "User Guide" => [joinpath("user_guide", f) for f in 
-            ("scm.md", "estimands.md", "estimation.md", "misc.md")],
+            ("scm.md", "estimands.md", "estimation.md")],
         "Examples" => [
             joinpath("examples", "super_learning.md"),
             joinpath("examples", "double_robustness.md")
             ],
+        "Integrations" => "integrations.md",
         "Estimators' Cheat Sheet" => "estimators_cheatsheet.md",
-        "Resources" => "resources.md",
+        "Learning Resources" => "resources.md",
         "API Reference" => "api.md",
-        "Integrations" => "integrations.md",
+
     ],
     pagesonly=true,
     clean = true,

docs/src/user_guide/estimation.md

@@ -167,7 +167,7 @@ There are some practical considerations
 
 - Choice of `resampling` Strategy: The theory behind sample-splitting requires the nuisance functions to be sufficiently well estimated on **each and every** fold. A practical aspect of it is that each fold should contain a sample representative of the dataset. In particular, when the treatment and outcome variables are categorical it is important to make sure the proportions are preserved. This is typically done using `StratifiedCV`.
 - Computational Complexity: Sample-splitting results in ``K`` fits of the nuisance functions, drastically increasing computational complexity. In particular, if the nuisance functions are estimated using (P-fold) Super-Learning, this will result in two nested cross-validation loops and ``K \times P`` fits.
-- Caching of Nuisance Functions: Because the `resampling` strategy typically needs to preserve the outcome and treatment proportions, very little reuse of cached models is possible (see [Caching Models](@ref)).
+- Caching of Nuisance Functions: Because the `resampling` strategy typically needs to preserve the outcome and treatment proportions, very little reuse of cached models is possible (see [Using the Cache](@ref)).
 
 ## Using the Cache
 

src/cache.jl

@@ -3,8 +3,23 @@ MLJBase.report(factors::MLCMRelevantFactors) = MLJBase.report(factors.outcome_me
 
 MLJBase.report(cache) = MLJBase.report(cache[:targeted_factors])
 
+"""
+    gradients(cache)
+
+Retrieves the gradients corresponding to each targeting step from the cache.
+"""
 gradients(cache) = MLJBase.report(cache[:targeted_factors]).gradients
 
+"""
+    estimates(cache)
+
+Retrieves the estimates corresponding to each targeting step from the cache.
+"""
 estimates(cache) = MLJBase.report(cache[:targeted_factors]).estimates
 
+"""
+    epsilons(cache)
+
+Retrieves the fluctuations' epsilons corresponding to each targeting step from the cache.
+"""
 epsilons(cache) = MLJBase.report(cache[:targeted_factors]).epsilons