More documentation fixes

src-full/cls/omlf_logistic_regression.mli

@@ -18,7 +18,7 @@
 (** Use
   {{:https://en.wikipedia.org/wiki/Logistic_regression}Logistic Regression}
   to estimate log-odds for each of the quantitative features in the
-  {{!modtype:Cls_intf.Continuous_encoded_data}encoded data},
+  {{!modtype:Oml.Classification.Input_interfaces.Continuous_encoded_data}encoded data},
   per class.
 
   Unless specified via the [classes] argument to [estimate], the first
@@ -29,7 +29,7 @@
 
   A constant [1] is added to all encoded features by [estimate],
   there is no need to add one with
-  {{!val:Cls_intf.Continuous_encoded_data.encoding}encoding}. *)
+  {{!val:Oml.Classification.Input_interfaces.Continuous_encoded_data.encoding}encoding}. *)
 module Binary(D: Oml.Classification.Input_interfaces.Continuous_encoded_data) :
   sig
     include Oml.Classification.Classifier_interfaces.Classifier
@@ -58,16 +58,13 @@ module Binary(D: Oml.Classification.Input_interfaces.Continuous_encoded_data) :
     val base_class : t -> class_
   end
 
-(** Use
-  {{:https://en.wikipedia.org/wiki/Multinomial_logistic_regression}
+(** Use {{:https://en.wikipedia.org/wiki/Multinomial_logistic_regression}
   Multiple Class Logistic Regression} to estimate log-odds
   for each of the quantitative features in the
-  {{!modtype:Cls_intf.Continuous_encoded_data}encoded data},
-  per class.
-
-  A constant [1] is added to all encoded features by [estimate],
+  {{!modtype:Oml.Classification.Input_interfaces.Continuous_encoded_data}encoded data},
+  per class: A constant [1] is added to all encoded features by [estimate],
   there is no need to add one with
-  {{!val:Cls_intf.Continuous_encoded_data.encoding}encoding}. *)
+  {{!val:Oml.Classification.Input_interfaces.Continuous_encoded_data.encoding}encoding}. *)
 module Multiclass(D: Oml.Classification.Input_interfaces.Continuous_encoded_data) :
   sig
     include Oml.Classification.Classifier_interfaces.Classifier

src-full/cls/omlf_naive_bayes.mli

@@ -21,7 +21,7 @@ include module type of Oml.Classification.Naive_bayes
   {{:https://en.wikipedia.org/wiki/Naive_Bayes_classifier#Gaussian_naive_Bayes}
   Gaussian Naive Bayes} by estimating mean and standard deviations
   for each of the quantitative features in the
-  {{!modtype:Cls_intf.Continuous_encoded_data}encoded data}. *)
+  {{!modtype:Oml.Classification.Input_interfaces.Continuous_encoded_data}encoded data}. *)
 module Gaussian(D: Oml.Classification.Input_interfaces.Continuous_encoded_data) : sig
   include Oml.Classification.Classifier_interfaces.Generative
       with type feature := D.feature

src/oml.mli

@@ -64,8 +64,8 @@ end
 
 (** Inference, parameter estimation.*)
 module Statistics : sig
-  module Continued_fraction : sig include
-    module type of Oml_continued_fraction
+  module Continued_fraction : sig
+    include module type of Oml_continued_fraction
   end
   module Functions : sig include module type of Oml_functions end
   module Sampling : sig include module type of Oml_sampling end
@@ -82,14 +82,14 @@ module Classification : sig
   module Input_interfaces : sig include module type of Oml_classification_input_interfaces end
   module Classifier_interfaces : sig
     (** A {{!type:t}classifier}, once {{!val:estimate} estimated} from
-        the {{!type:Data.feature}features} found in
+        the {{!type:Input_interfaces.Data.feature}features} found in
         {{!type:samples}samples} of data, assigns
         {{!Probabilities.t}probabilities} to
-        {{!type:Data.class_}classes}
+        {{!type:Input_interfaces.Data.class_}classes}
         on future samples when {{!val:eval}evaluated}. *)
     module type Classifier = sig
       include Input_interfaces.Data
-      include Oml_util.Optional_arg_intf
+      include Util.Optional_arg_intf
 
       (** The classifier. *)
       type t

src/stats/oml_descriptive.mli

@@ -17,12 +17,6 @@
 
 (** Compute statistics (aka estimators) and other methods
     {!val:Descriptive.histogram} that describe data stored in a float array.
-
-    Testing against statistics is found in {!module:Hypothesis_testing}. These
-    algorithms are not necessarily tailored for the most accurate algorithms
-    that try to minimize truncating (see {!module:Online} for online
-    algorithms) but for having a simple function(al) interface with good runtime
-    performance.
 *)
 
 (* Central tendency *)