final

main.jl

@@ -1,11 +1,19 @@
-using ScikitLearn
-using ScikitLearn.CrossValidation: train_test_split
-using PyCall
-np = pyimport("numpy")
+using Random
+
+function train_test_split(data, y_column, train_size)
+    #set seed = 30 -> fix random state
+    Random.seed!(30);
+    n = size(data,1)
+    #random
+    idx = shuffle(Vector(1:n))
+    #split train and test dataset
+    train_idx = view(idx, 1:floor(Int, train_size*n))
+    test_idx = view(idx, (floor(Int, train_size*n)+1):n)
+    data[train_idx, :], y_column[train_idx, :],  data[test_idx, :], y_column[test_idx, :]
+end
 
 
 function read_file(filename)
-    # X = Array{Float64}(undef, 0,2)
     X=[]
     y = []
     fp = open(filename,"r")
@@ -17,20 +25,19 @@ function read_file(filename)
         popfirst!(x_line)
         y_line = pop!(x_line)
         x_line = [parse(Float64,ss) for ss in x_line]
-        # x_line = np.asarray(x_line).astype("float64")
+
+        #create x
         append!(X, [x_line])
-        # X = [X;x_line]
+
+        #create y
         push!(y, y_line)
     end
-    # X= np.asarray(X)
-    # y = np.asarray(y)
-
     X = hcat(X...)'
     close(fp)
 
-    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=1/3, random_state=42);
+    # X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=1/3, random_state=42);
+    X_train, y_train, X_test, y_test = train_test_split(X, y,2/3)
     return X_train, y_train, X_test, y_test, attr
-    # return X, y, attr
 end
 
 mutable struct Node
@@ -65,9 +72,6 @@ mutable struct DecisionTree
         tree = new(min_value_gain,depth,attr)
         return tree
     end
-
-
-
 end
 
 function fit(X_train, y_train, tree, attr)
@@ -86,10 +90,11 @@ function fit(X_train, y_train, tree, attr)
     #init tree
     tree.root = root
     list_root_check = [root]
+    #process with tree's child
     while length(list_root_check) != 0
         active_node = pop!(list_root_check)
         if active_node.entropy < tree.min_value_gain || active_node.depth < tree.depth
-            active_node.child = make_split(active_node, X_train, y_train, attr)
+            active_node.child = make_split(active_node, X_train, y_train)
             if length(active_node.child) == 0
                 y_dummy = [y_train[i] for i in active_node.pos]
                 vals_unique = unique(y_dummy)
@@ -112,7 +117,7 @@ function fit(X_train, y_train, tree, attr)
     return tree
 end
 
-function make_split(node::Node, x_train, y_train, attr)
+function make_split(node::Node, x_train, y_train)
     choose_gain = 0
     choose_split = []
     choose_attr = ""
@@ -122,12 +127,13 @@ function make_split(node::Node, x_train, y_train, attr)
     x_train_T = copy(x_train)'
     pos = node.pos
     for col in 1:length(x_train_T[:,1])
-        H_min = 9999
+        H_min = 10
         left_set_choose = []
         right_set_choose = []
         entropy_list_choose = []
         value_choose = 0
 
+        #get unique value to choose cutoff
         unique_value = Set(sort(x_train_T[col,:]))
         if length(unique_value) == 1
             continue
@@ -162,6 +168,7 @@ function make_split(node::Node, x_train, y_train, attr)
                     print("x")
                 end
             end
+            #choose cutoff for each column
             if H < H_min 
                 H_min = H
                 left_set_choose = left_set
@@ -173,7 +180,10 @@ function make_split(node::Node, x_train, y_train, attr)
         if minimum([length(left_set_choose), length(right_set_choose)]) < 2
             continue
         end
+
+        #calc information gain
         gain_information = node.entropy - H_min
+        #choose column which have a higher information gain
         if gain_information > choose_gain
             choose_gain = gain_information
             choose_attr = col
@@ -185,6 +195,7 @@ function make_split(node::Node, x_train, y_train, attr)
     node.name = string(choose_attr)
     node.value_split = best_value
     k = 1
+    #create new child
     for split in choose_split
         if k == 1
             new_node = Node("< "*string(best_value),entropy_best[k],split,node.depth+1)
@@ -195,7 +206,6 @@ function make_split(node::Node, x_train, y_train, attr)
         end
         k += 1
     end
-    println("1")
     return child_nodes
 end
 
@@ -227,18 +237,17 @@ function accuracy(y_predict, y_true)
 end
 
 function main()
-    X_train, y_train, X_test, y_test, attr = read_file("Iris.csv")
-    # X_train, y_train, attr = read_file("Iris.csv")
+    X_train, y_train, X_test, y_test, attr = read_file(joinpath(@__DIR__,"Iris.csv"))
     tree = DecisionTree(0, 10, attr)
     fit(X_train, y_train, tree, attr)
-    y_hat_train = predict(tree,X_train)
+    y_hat_train = predict(tree, X_train)
     acc_train = accuracy(y_hat_train, y_train)
 
-    y_hat_test = predict(tree,X_test)
+    y_hat_test = predict(tree, X_test)
     acc_test = accuracy(y_hat_test, y_test)
 
-    println("Accuracy of train dataset: ",acc_train)
-    println("Accuracy of test dataset: ",acc_test)
+    println("Accuracy of train dataset: ",acc_train,"\n")
+    println("Accuracy of test dataset: ",acc_test,"\n")
 end
 
 main()