Jupyter notebook for MLJ using Iris data and DecisionTree

iris.ipynb

iris_v2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[34mSupervisedTask{} @ 3…62\u001b[39m\n"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using MLJ\n",
    "iris = load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[34mSupervisedTask{} @ 3…62\u001b[39m\n"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "import MLJModels ✔\n",
      "import DecisionTree ✔\n",
      "import MLJModels.DecisionTree_.DecisionTreeClassifier ✔\n"
     ]
    }
   ],
   "source": [
    "@load DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(pruning_purity = 1.0,\n",
       "                       max_depth = 2,\n",
       "                       min_samples_leaf = 1,\n",
       "                       min_samples_split = 2,\n",
       "                       min_purity_increase = 0.0,\n",
       "                       n_subfeatures = 0.0,\n",
       "                       display_depth = 5,\n",
       "                       post_prune = false,\n",
       "                       merge_purity_threshold = 0.9,)\u001b[34m @ 8…98\u001b[39m"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_model = DecisionTreeClassifier(max_depth=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[34mMachine{DecisionTreeClassifier} @ 8…47\u001b[39m\n"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree = machine(tree_model, iris)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Info: Training \u001b[34mMachine{DecisionTreeClassifier} @ 8…47\u001b[39m.\n",
      "└ @ MLJ /home/simon/.julia/dev/MLJ/src/machines.jl:130\n"
     ]
    },
    {
     "ename": "MethodError",
     "evalue": "MethodError: no method matching build_tree(::CategoricalArrays.CategoricalArray{String,1,UInt32,String,CategoricalArrays.CategoricalString{UInt32},Union{}}, ::Array{Float64,2}, ::Float64, ::Int64, ::Int64, ::Int64, ::Float64)\nClosest candidates are:\n  build_tree(!Matched::Array{T<:Float64,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any, ::Any; rng) where {S, T<:Float64} at /home/simon/.julia/dev/DecisionTree/src/regression/main.jl:27\n  build_tree(!Matched::Array{T,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any, ::Any; rng) where {S, T} at /home/simon/.julia/dev/DecisionTree/src/classification/main.jl:83\n  build_tree(!Matched::Array{T<:Float64,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any) where {S, T<:Float64} at /home/simon/.julia/dev/DecisionTree/src/regression/main.jl:27\n  ...",
     "output_type": "error",
     "traceback": [
      "MethodError: no method matching build_tree(::CategoricalArrays.CategoricalArray{String,1,UInt32,String,CategoricalArrays.CategoricalString{UInt32},Union{}}, ::Array{Float64,2}, ::Float64, ::Int64, ::Int64, ::Int64, ::Float64)\nClosest candidates are:\n  build_tree(!Matched::Array{T<:Float64,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any, ::Any; rng) where {S, T<:Float64} at /home/simon/.julia/dev/DecisionTree/src/regression/main.jl:27\n  build_tree(!Matched::Array{T,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any, ::Any; rng) where {S, T} at /home/simon/.julia/dev/DecisionTree/src/classification/main.jl:83\n  build_tree(!Matched::Array{T<:Float64,1}, ::Array{S,2}, ::Any, ::Any, ::Any, ::Any) where {S, T<:Float64} at /home/simon/.julia/dev/DecisionTree/src/regression/main.jl:27\n  ...",
      "",
      "Stacktrace:",
      " [1] fit(::DecisionTreeClassifier, ::Int64, ::DataFrames.DataFrame, ::CategoricalArrays.CategoricalArray{String,1,UInt32,String,CategoricalArrays.CategoricalString{UInt32},Union{}}) at /home/simon/.julia/dev/MLJModels/src/DecisionTree.jl:110",
      " [2] #fit!#3(::Array{Int64,1}, ::Int64, ::Bool, ::Function, ::Machine{DecisionTreeClassifier}) at /home/simon/.julia/dev/MLJ/src/machines.jl:131",
      " [3] (::getfield(StatsBase, Symbol(\"#kw##fit!\")))(::NamedTuple{(:rows,),Tuple{Array{Int64,1}}}, ::typeof(fit!), ::Machine{DecisionTreeClassifier}) at ./none:0",
      " [4] top-level scope at In[6]:2"
     ]
    }
   ],
   "source": [
    "train, test = partition(eachindex(iris), 0.7, shuffle=true); # 70:30 split\n",
    "fit!(tree, rows=train)\n",
    "yhat = predict(tree, iris);\n",
    "misclassification_rate(yhat, iris)"
   ]
  }
 ],
 "metadata": {
  "@webio": {
   "lastCommId": null,
   "lastKernelId": null
  },
  "kernelspec": {
   "display_name": "Julia 1.1.0",
   "language": "julia",
   "name": "julia-1.1"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.1.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}