new struct ModelEquation, best print of fitted regressions

docs/src/tutorial.md

@@ -74,17 +74,26 @@ data = DataFrame(
 
 # Performing the regression analysis on the full dataset
 # Here, we analyze the relationship between height (h) and diameter (dbh)
-models = regression(:h, :dbh, data)
+models = regression(:h, :dbh, data);
+# Alternative print of fitted models
+models_eq = ModelEquation.(models)
+```
 
+```@example regression_data
+#number of fitted regressions
+length(models)
 ```
 
 ### Viewing the top models based on specific criteria
 
 ```@example regression_data
-best_models = criteria_table(models, :adjr2, :rmse)
+# Using all criteria and presenting the 10 best models
+best_models = criteria_table(models)
+```
 
-# Chossing as especific criteria
-best_models = criteria_table(models, :adjr2, :rmse)
+```@example regression_data
+# Chossing as especific criteria and best 5, if best = false or 0 show all regressions from 'models'
+best_models_v2 = criteria_table(models, :adjr2, :rmse, best=5)
 ```
 
 ### Regression Selection Criteria [`criteria_table`](@ref)

src/ForestMensuration.jl

@@ -53,6 +53,7 @@ include("show.jl")
 
 export
   # Regression
+  ModelEquation,
   TableRegressionModel,
   regression,
   prediction,

src/criteria_functions.jl

@@ -190,13 +190,15 @@ function criteria_table(
 
   # If 'best' is false, return the full DataFrame
   if best === false
+    ct.model = ModelEquation.(ct.model)
     return ct
   elseif best < length(model)
     # If 'best' is less than the number of models, return the top 'best' models
     top_models = ct[1:best, 1]
     return criteria_table(top_models, criteria...; best=false) # Re-run with selected models
   else
     # Otherwise, return the full DataFrame
+    ct.model = ModelEquation.(ct.model)
     return ct
   end
 end

src/show.jl

@@ -1,5 +1,10 @@
 import Base: show
 
+# Define the show method for custom display of ModelEquation
+function show(io::IO, model::ModelEquation)
+  print(io, model.output)
+end
+
 # Custom show method for SiteAnalysis to display  the site_table and site_plot
 function show(io::IO, analysis::SiteAnalysis)
   show(io, analysis.site_table)

src/structs_consts.jl

@@ -87,6 +87,36 @@ Newton Method:
 """
 abstract type Newton <: CubingMethod end
 
+
+"""
+    struct ModelEquation
+
+Define ModelEquation struct to store the regression results
+  # Fields
+  - output::String
+  - model::TableRegressionModel
+"""
+struct ModelEquation
+  output::String
+  model::TableRegressionModel
+  # Inner constructor to initialize `output` based on `model`
+  function ModelEquation(model::TableRegressionModel)
+    # Get coefficients and terms from the model
+    β = coef(model)
+    n = length(β)
+    output = string(StatsModels.coefnames(model.mf.f.lhs)) * " = $(round(β[1], digits = 6))"
+
+    for i in 2:n
+      term = coefnames(model)[i]
+      product = string(round(abs(β[i]), sigdigits=6)) * " * " * term
+      output *= signbit(β[i]) ? " - $(product)" : " + $(product)"
+    end
+
+    # Return the constructed RegressionEquation object
+    new(output, model)
+  end
+end
+
 """
     struct SiteAnalysis