onehotencodingTest.cpp

Go to the documentation of this file.

//gcc -I/Users/xinyuzhang/Desktop/Spring2021/c++pattern/project/cosan /Users/xinyuzhang/Desktop/Spring2021/c++pattern/project/cosan/test/selection/GridSearchTest.cpp
 
#include <iostream>
#include <vector>
#include <string>
#include <cosan/data/CosanData.h>
// #include <cosan/model/CosanLinearModel.h>
#include <cosan/model/CosanLinearRegression.h>
#include <cosan/model/CosanRidgeRegression.h>
#include <cosan/model/CosanPCRRidge.h>
#include <cosan/selection/gridsearch.h>
#include <cosan/selection/randomgridsearch.h>
#include <cosan/evaluation/CosanMetric.h>
#include <cosan/selection/kfold.h>
#include <cosan/selection/randomkfold.h>
#include <cosan/selection/timeseriessplit.h>
#include <cosan/preprocessing/encoder.h>
 
 
typedef double NumericType;
using namespace Cosan;
int main(){
    std::vector<std::vector<std::string>> svalues;
    std::vector<std::unordered_map<std::string, gsl::index>> categories;    
    std::vector<std::string>  svaluesX = std::vector<std::string>({"a","b","c","b","a","c"});
    gsl::index colCat = 3 ;
    svalues = std::vector<std::vector<std::string>>(colCat,std::vector<std::string>(svaluesX.size()/colCat,"") );
    CosanMatrix<NumericType> OneHotMatrixi;
    CosanMatrix<NumericType> newMatrix;
    for (gsl::index col = 0;col<colCat;col++){
        std::unordered_map<std::string, gsl::index> categoryToOrdinal{};
        for (gsl::index row = 0;row<svaluesX.size()/colCat;row++){
            svalues[col][row] = svaluesX[row*colCat+col];
            if (categoryToOrdinal.find(svalues[col][row]) == categoryToOrdinal.end()) {
                categoryToOrdinal.insert({svalues[col][row], categoryToOrdinal.size()});
            }
        }
        categories.push_back(categoryToOrdinal);
        OneHotMatrixi = CosanMatrix<NumericType>::Zero(svaluesX.size()/colCat,categoryToOrdinal.size());
        for (gsl::index row = 0;row<svaluesX.size()/colCat;row++){
            OneHotMatrixi(row,categoryToOrdinal[svalues[col][row]]) = 1;
        }
        if (col ==0){
            newMatrix =OneHotMatrixi;
        }
        else{
            for (gsl::index i = 0;i<OneHotMatrixi.cols();i++){
                newMatrix.conservativeResize(newMatrix.rows(), newMatrix.cols()+1);
                newMatrix.col(newMatrix.cols()-1) = OneHotMatrixi.col(i);
            }
        }
 
    }
 
    // std::vector<std::string>  svaluesX = std::vector<std::string>({"a","b","c","b","a","c"});
    // gsl::index colCat = 2;
    // std::vector<std::unordered_map<std::string, gsl::index>> categories;
    // std::vector<std::vector<std::string>> svalues;
    // svalues = std::vector<std::vector<std::string>>(colCat,std::vector<std::string>(svaluesX.size()/colCat,"") );
    // Cosan::CosanMatrix<double> OneHotMatrixi;
    // Cosan::CosanMatrix<double> newMatrix;
    // for (gsl::index col = 0;col<colCat;col++){
    //     std::unordered_map<std::string, gsl::index> categoryToOrdinal{};
    //     for (gsl::index row = 0;row<svaluesX.size()/colCat;row++){
    //         svalues[col][row] = svaluesX[row*colCat+col];
    //         if (categoryToOrdinal.find(svalues[col][row]) == categoryToOrdinal.end()) {
    //             categoryToOrdinal.insert({svalues[col][row], categoryToOrdinal.size()});
    //         }
    //     }
    //     categories.push_back(categoryToOrdinal);
    //     OneHotMatrixi = Cosan::CosanMatrix<double>::Zero(svaluesX.size()/colCat,categoryToOrdinal.size());
 
    //     for (gsl::index row = 0;row<svaluesX.size()/colCat;row++){
    //         OneHotMatrixi(row,categoryToOrdinal[svalues[col][row]]) = 1;
    //     }
    //     if (col ==0){
    //         newMatrix =OneHotMatrixi;
    //     }
    //     else{
    //         for (gsl::index i = 0;i<OneHotMatrixi.cols();i++){
    //             newMatrix.conservativeResize(newMatrix.rows(), newMatrix.cols()+1);
    //             newMatrix.col(newMatrix.cols()-1) = OneHotMatrixi.col(i);
    //         }
    //     }
 
    // }
    std::cout<<newMatrix<<std::endl;
 
    // int x = 5;
    // int y = 5;
    // Cosan::CosanMatrix<double> X_input = Cosan::CosanMatrix<double>::Zero(x,y);
    // // std::cout<<X_input.rows()<<std::endl;
    // // X_input.setZero();
    // std::cout<<X_input<<std::endl;   
    // Cosan::CosanMatrix<db> Y_input;  
    // constexpr gsl::index nrows = 10000;
    // constexpr gsl::index ncols = 5;
    // X_input.resize(nrows,ncols);
    // Y_input.resize(nrows,1);
    // // X_input = Eigen::Matrix<decltype(X_input)::Scalar,nrows,ncols>::Random();
    // // Y_input = Eigen::Matrix<decltype(X_input)::Scalar,nrows,1>::Random();
    // // std::cout<<X_input <<std::endl;
    // // std::cout<<Y_input <<std::endl;
    // // X_input = Eigen::Matrix<X_input::Scalr,X_input.rows(),X_input.cols()>::Random();  
 
    // // Cosan::CosanData<db>  CRD(X_input,Y_input);
    // Cosan::CosanData<db>  CRD(Eigen::Matrix<decltype(X_input)::Scalar,nrows,ncols>::Random()
    //                          ,Eigen::Matrix<decltype(X_input)::Scalar,nrows,1>::Random());
    // // Cosan::CosanLinearRegression<decltype(X_input)::Scalar> CRRwbias(true);
    // Cosan::CosanPCRRidge<decltype(X_input)::Scalar> CRRwbias;
 
    // CRRwbias.fit(CRD.GetInput(),CRD.GetTarget());
    // // std::cout<<CRRwbias.GetParams()<<std::endl;
    // Cosan::CosanMatrix<db> beta = CRRwbias.GetBeta();
    // std::cout<<CRRwbias.predict(CRD.GetInput())<<std::endl;
    // std::cout<<beta<<std::endl;
    // Cosan::MeanSquareError<decltype(X_input)::Scalar> mse;
    // Cosan::MeanAbsError<decltype(X_input)::Scalar> mae;
    // Cosan::R2Score<decltype(X_input)::Scalar> r2;
    // Cosan::MaxError<decltype(X_input)::Scalar> me;
 
 //    db a = 0.05;
 //    std::vector<db> v(10);
 //    std::generate(v.begin(), v.end(), [n = 1, &a]() mutable { return n++ * a; });
 //    Cosan::RandomKFold kf(5);
 //    for (auto & each : v){
 //     std::cout<<each<<std::endl;
 //    }
 
 //    // KFoldParallel
 //    // TimeSeriesSplitParallel
 //    // TimeSeriesSplit
 //    // RandomKFoldParallel
 //    // RandomKFold
    // // Cosan::RandomGridSearch GDS1(CRD,CRRwbias,mse,kf,v);
    // std::cout<<Cosan::RandomGridSearch(CRD,CRRwbias,mse,kf,v).GetBestParams()<<std::endl;    
    // // Cosan::RandomGridSearch GDS2(CRD,CRRwbias,mae,kf,v);
    // std::cout<<GDS2.GetBestParams()<<std::endl;
    // Cosan::RandomGridSearch GDS3(CRD,CRRwbias,r2,kf,v);
    // std::cout<<GDS3.GetBestParams()<<std::endl;  
    // Cosan::RandomGridSearch GDS4(CRD,CRRwbias,me,kf,v);
    // std::cout<<GDS4.GetBestParams()<<std::endl;  
 
 
    // std::cout<<GDS.GetBestParams()<<std::endl;
   // std::cout<<CRRwbias.predict(X_input)<<std::endl;
 
 
    return 0;
}