administrator/pascalabcnet
1
0
Fork 0
Code Issues Pull requests Actions Packages Projects Releases Wiki Activity

LeafValue для TreeNode

Browse source
This commit is contained in:
Mikhalkovich Stanislav 2026-02-19 09:43:42 +03:00
parent 31aaf0d13d
commit 5ddc24e36a
1 changed files with 12 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
bin/Lib/MLModelsABC.pas
View file

@ -238,8 +238,7 @@ type
Threshold: real;
Left: DecisionTreeNode;
Right: DecisionTreeNode;
ClassValue: integer;
Value: real;
LeafValue: real;
function Clone: DecisionTreeNode;
end;
@ -258,20 +257,11 @@ type
fMinSamplesLeaf: integer;
fFitted: boolean;
function BuildTree(
X: Matrix; y: Vector;
indices: array of integer;
depth: integer
): DecisionTreeNode;
function BuildTree(X: Matrix; y: Vector; indices: array of integer; depth: integer): DecisionTreeNode;
function FindBestSplit(
X: Matrix; y: Vector;
indices: array of integer
): SplitResult;
function FindBestSplit(X: Matrix; y: Vector; indices: array of integer): SplitResult;
function LeafValue(
y: Vector; indices: array of integer
): real; virtual; abstract;
function LeafValue(y: Vector; indices: array of integer): real; virtual; abstract;
function LeafNode(value: real): DecisionTreeNode;
@ -279,11 +269,8 @@ type
procedure CopyBaseState(dest: DecisionTreeBase);
public
constructor Create(
maxDepth: integer := 10;
minSamplesSplit: integer := 2;
minSamplesLeaf: integer := 1
);
constructor Create(maxDepth: integer := 10; minSamplesSplit: integer := 2;
minSamplesLeaf: integer := 1);
function IsFitted: boolean;
end;
@ -1239,7 +1226,7 @@ function LeafClass(c: integer): DecisionTreeNode;
begin
var n := new DecisionTreeNode;
n.IsLeaf := true;
n.ClassValue := c;
n.LeafValue := c;
Result := n;
end;
@ -1247,7 +1234,7 @@ function LeafValue(v: real): DecisionTreeNode;
begin
var n := new DecisionTreeNode;
n.IsLeaf := true;
n.Value := v;
n.LeafValue := v;
Result := n;
end;
@ -1270,8 +1257,7 @@ begin
n.IsLeaf := IsLeaf;
n.FeatureIndex := FeatureIndex;
n.Threshold := Threshold;
n.ClassValue := ClassValue;
n.Value := Value;
n.LeafValue := LeafValue;
if Left <> nil then
n.Left := Left.Clone;
@ -1314,7 +1300,7 @@ function DecisionTreeBase.LeafNode(value: real): DecisionTreeNode;
begin
var n := new DecisionTreeNode;
n.IsLeaf := true;
n.Value := value;
n.LeafValue := value;
Result := n;
end;
@ -1464,7 +1450,7 @@ begin
node := node.Right;
end;
Result := integer(node.Value); // internal class index
Result := integer(node.LeafValue); // internal class index
end;
@ -1680,7 +1666,7 @@ begin
node := node.Right;
end;
Result := node.Value;
Result := node.LeafValue;
end;
function DecisionTreeRegressor.Predict(X: Matrix): Vector;