easy_ml 0.2.0.pre.rc84 → 0.2.0.pre.rc88

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ff472f1d00b1629047b7b414ae6d8104bc34529602808d919e5117ba4fd99077
-  data.tar.gz: 676ba30c99251193848aa2f02957e18d1abc9fdfef112ab0699720e8edc96d1a
+  metadata.gz: f9a3eb82ce50657d230c9d00a75f84f7f09ce96b68d58fa8a1d615ad4fd31d77
+  data.tar.gz: f0ded08a9c73232bcea3eb23b9e678bc392647873e53c425c6b21f62104c983b
 SHA512:
-  metadata.gz: a69606a0cf1a10939c5fa29ac8425e2cfbec08dc99e4c0dda510b7cc227171654a89f11ac9cba987636e5cf2fa7ada7d98e179dc04f880514fbf6013f2f31bda
-  data.tar.gz: 5bf4737b1d4fa4dcefa999be84e76152c3a665826a5e55d72a9a120fbf5dd79cf53a44a4f38dc78fa6abc40f38e1c13779c0ab7b98f219fe29c29005fd1e5134
+  metadata.gz: 76b6f960c24f0032d4959f7435770f64245a5699f97150118127fac618e46965267ae996c3be3b4bb21a0734177c039ec2357b690a69ede3117c834e3ff09083
+  data.tar.gz: 241ae30e3e2934eef6a74aae16aa05cfb1e10ed9145562045ccb93403c292b46449003badee5a0798a70987394469d6781b6ef8d14d5a552ee9da954efc75f44

data/app/controllers/easy_ml/datasets_controller.rb

@@ -23,7 +23,7 @@
 module EasyML
   class DatasetsController < ApplicationController
     def index
-      datasets = Dataset.all.includes(:columns, :datasource).order(id: :desc)
+      datasets = Dataset.all.includes(:datasource, columns: [:lineages]).order(id: :desc)
 
       render inertia: "pages/DatasetsPage", props: {
         datasets: datasets.map { |dataset| dataset_to_json_small(dataset) },
@@ -73,7 +73,23 @@ module EasyML
 
       # Iterate over columns to check and update preprocessing_steps
       dataset_params[:columns_attributes]&.each do |_, column_attrs|
-        column_attrs[:preprocessing_steps] = nil if column_attrs.dig(:preprocessing_steps, :training, :method) == "none"
+        if column_attrs.dig(:preprocessing_steps, :training, :method) == "none"
+          column_attrs[:preprocessing_steps] = nil
+        elsif column_attrs.dig(:preprocessing_steps, :training)
+          # Ensure encoding is properly set for categorical columns
+          training_config = column_attrs.dig(:preprocessing_steps, :training)
+          if training_config[:params]
+            # Remove old encoding params as they're now part of the encoding field
+            training_config[:params].delete(:one_hot)
+            training_config[:params].delete(:ordinal_encoding)
+          end
+
+          # Ensure embedding params are present when encoding is embedding
+          if training_config[:encoding] == "embedding" && training_config[:params]
+            training_config[:params][:llm] ||= "openai"
+            training_config[:params][:model] ||= "text-embedding-3-small"
+          end
+        end
       end
 
       # Handle feature ID assignment for existing features
@@ -165,7 +181,7 @@ module EasyML
     private
 
     def preprocessing_params
-      [:method, { params: [:constant, :categorical_min, :one_hot, :ordinal_encoding, { clip: %i[min max] }] }]
+      [:method, :encoding, { params: [:constant, :categorical_min, :llm, :model, :preset, :dimensions, { clip: %i[min max] }] }]
     end
 
     def dataset_params