Raw Data container. More...

#include <CosanData.h>

Inheritance diagram for Cosan::CosanRawData< NumericType >:

Public Member Functions
	CosanRawData ()=default

	CosanRawData (const std::string &srcX, const std::string &srcY)
	Constructor: Read data X and Y from csv files and form raw data container. More...

	CosanRawData (const std::string &srcX)
	Constructor: Read data X from csv and form raw data container. More...

void	SetInput (const std::string &srcX)
	Update input X from csv file. More...

void	SetTarget (const std::string &srcY)
	Update target Y from csv file. More...

void	ConcatenateData (const CosanMatrix< NumericType > &inputX)
	Concatenate X using CosanMatrix<NumericType> input X. Add new columns. More...

void	UpdateData (const CosanMatrix< NumericType > &inputX)
	Update X using CosanMatrix<NumericType> input X. More...

void	UpdateData (const CosanMatrix< NumericType > &inputX, const CosanMatrix< NumericType > &inputY)
	Update X and Y using CosanMatrix<NumericType> inputX,inputY. More...

void	UpdateCat (const std::vector< std::string > &inputX)
	Update categorical vector svaluesX using std::vector<std::string> & inputX. More...

void	UpdateCat (const std::vector< std::string > &inputX, const std::vector< std::string > &inputY)
	Update categorical vector svaluesX,svaluesY using std::vector<std::string> & inputX,inputY. More...

CosanMatrix< NumericType >	GetInput ()
	Get a copy of CosanMatrix<NumericType> X. More...

CosanMatrix< NumericType >	GetTarget ()
	Get a copy of CosanMatrix<NumericType> Y. More...

const CosanMatrix< NumericType > &	GetInput () const
	Get a const reference to const CosanMatrix<NumericType> X. More...

const CosanMatrix< NumericType > &	GetTarget () const
	Get a const reference to const CosanMatrix<NumericType> Y. More...

std::tuple< gsl::index, gsl::index >	GetMissingNumber ()
	Get the total number data information. More...

virtual const std::string	GetName () const
	Get the name of the objects. More...

const std::string &	GetSummaryMessageX () const
	Get the summary message on reading csv file on X. More...

const std::string &	GetSummaryMessageY () const
	Get the summary message on reading csv file on Y. More...

std::unordered_map< gsl::index, gsl::index > &	GetRawToNumIdx ()
	Raw data column index to numeric data matrix X column index. More...

std::unordered_map< gsl::index, gsl::index > &	GetRawToCatIdx ()
	Raw data column index to categorical data column index. More...

std::vector< std::vector< gsl::index > >	GetIdxpinfX () const
	Get the position of positive infinity in the origin data X. More...

std::vector< std::vector< gsl::index > >	GetIdxminfX () const
	Get the position of negative infinity in the origin data X. More...

std::vector< std::vector< gsl::index > >	GetIdxmissingX () const
	Get the position of missing in the origin data X. More...

std::vector< std::vector< gsl::index > >	GetIdxpinfY () const
	Get the position of positive infinity in the origin data Y. More...

std::vector< std::vector< gsl::index > >	GetIdxminfY () const
	Get the position of negative infinity in the origin data Y. More...

std::vector< std::vector< gsl::index > >	GetIdxmissingY () const
	Get the position of missing in the origin data Y. More...

std::set< gsl::index >	GetcolCatX () const
	Get the column index (in the origin X of csv file) where the column is of categorical type. More...

std::set< gsl::index >	GetcolCatY () const
	Get the column index (in the origin Y of csv file) where the column is of categorical type. More...

bool	GetcatY () const
	True if Y is categorical data type. False otherwise. More...

gsl::index	GetrowsX ()
	Get the number of rows for X. More...

gsl::index	GetrowsY ()
	Get the number of rows for Y. More...

gsl::index	GetcolsX ()
	Get the number of columns for X. More...

gsl::index	GetcolsY ()
	Get the number of columns for Y. More...

std::vector< std::string >	GetsvaluesX () const
	Get the vector of categorical data from X. order: row first. More...

std::vector< std::string >	GetsvaluesY () const
	Get the vector of categorical data from Y. order: row first. More...

CosanMatrix< NumericType >	GetType ()

Public Member Functions inherited from Cosan::CosanBO
	CosanBO ()
	Default constructor. More...

Protected Attributes
CosanMatrix< NumericType >	X
	Numeric data from origin CSV file for X. More...

CosanMatrix< NumericType >	Y
	Numeric data from origin CSV file for Y. More...

CosanMatrix< NumericType >	__TYPE

std::string	SummaryMessageX
	Loading message. More...

std::string	SummaryMessageY

std::vector< std::vector< gsl::index > >	IdxpinfX
	position for positive, negative infinity and missing values. More...

std::vector< std::vector< gsl::index > >	IdxminfX

std::vector< std::vector< gsl::index > >	IdxmissingX

std::vector< std::vector< gsl::index > >	IdxpinfY

std::vector< std::vector< gsl::index > >	IdxminfY

std::vector< std::vector< gsl::index > >	IdxmissingY

std::set< gsl::index >	colCatX
	column idx in the origin data that is categorical data. More...

std::set< gsl::index >	colCatY

bool	catY = false
	true mean respone variable Y is categorical data. More...

gsl::index	rowsX = 0
	number of rows. More...

gsl::index	colsX = 0

gsl::index	rowsY = 0
	number of columns More...

gsl::index	colsY = 0

std::vector< std::string >	svaluesX

std::vector< std::string >	svaluesY

Private Member Functions
std::tuple< gsl::index, gsl::index, std::string >	_load_csv (const std::string &path, CosanMatrix< NumericType > &X, std::vector< std::vector< gsl::index >> &Idxpinf, std::vector< std::vector< gsl::index >> &Idxminf, std::vector< std::vector< gsl::index >> &Idxmissing, std::vector< std::string > &svalues, std::set< gsl::index > &colCat)
	load data from csv More...

Private Attributes
std::unordered_map< gsl::index, gsl::index >	_raw2numIdx

std::unordered_map< gsl::index, gsl::index >	_raw2catIdx

Detailed Description

template<Numeric NumericType>
class Cosan::CosanRawData< NumericType >

Raw Data container.

Every constructor needs to have at least one input. To obtain CosanRawData, three constructors can be used:

CosanRawData(const std::string & srcX);

CosanRawData(const std::string & srcX, const std::string & srcY);

Definition at line 36 of file CosanData.h.

Constructor & Destructor Documentation

◆ CosanRawData() [1/3]

template<Numeric NumericType>

Cosan::CosanRawData< NumericType >::CosanRawData ( )

default

◆ CosanRawData() [2/3]

template<Numeric NumericType>

Cosan::CosanRawData< NumericType >::CosanRawData	(	const std::string &	srcX,
		const std::string &	srcY
	)

inline

Constructor: Read data X and Y from csv files and form raw data container.

Parameters

[in]	srcX	path to the csv file of data X;
[in]	srcY	path to the csv file of data Y.

Note: X and Y are from two separate csv files. For the data format of csv files, see Tutorial.

Definition at line 45 of file CosanData.h.

                                                                      :CosanBO(){
 //             static_assert(std::is_arithmetic<NumericType>::value, "NumericType must be numeric");
              if (std::is_same_v<NumericType, bool>){
                  throw std::invalid_argument(
                          "We do not accept bool at this moment. Try unsigned int, unsigned long, unsigned long long, int, "
                          "long, long, float, double ,long double.");
              }
              SetInput(srcX);
              SetTarget(srcY);
          }

◆ CosanRawData() [3/3]

template<Numeric NumericType>

Cosan::CosanRawData< NumericType >::CosanRawData ( const std::string & srcX )

inline

Constructor: Read data X from csv and form raw data container.

Parameters

[in] srcX path to the csv file of data X.

Definition at line 59 of file CosanData.h.

                                             :CosanBO(){
 //            static_assert(std::is_arithmetic<NumericType>::value, "NumericType must be numeric");
             if (std::is_same_v<NumericType, bool>){
                 throw std::invalid_argument(
                         "We do not accept bool at this moment. Try unsigned int, unsigned long, unsigned long long, int, "
                         "long, long, float, double ,long double.");
             }
             SetInput(srcX);}

Member Function Documentation

◆ _load_csv()

template<Numeric NumericType>

std::tuple<gsl::index,gsl::index,std::string> Cosan::CosanRawData< NumericType >::_load_csv	(	const std::string &	path,
		CosanMatrix< NumericType > &	X,
		std::vector< std::vector< gsl::index >> &	Idxpinf,
		std::vector< std::vector< gsl::index >> &	Idxminf,
		std::vector< std::vector< gsl::index >> &	Idxmissing,
		std::vector< std::string > &	svalues,
		std::set< gsl::index > &	colCat
	)

inlineprivate

load data from csv

We accept data file in csv format where each data is of dimension n\times p where n (the number of rows) is number of samples and p (the number of columns) denotes number of features. Each data entry is separated by "," and allows for positive/negative infinity (user-specific NumericType is float,double or long double), missing values (either emptry entry between two contiguous comma "," or NAN expression ) and non-number string. If user-specific NumericType is float,double or long double, acceptable numeric expressions also include hexadecimal and variants of decimal float-poing expression (see this for more details). It will throw std::invalid_argument if the the entry read is not-a-number expression except this entry is of categorical type.

We determine each column's data type (either numeric or categorical) by the first row. We treat every entry as numeric if it is a number (whether it is ordinal or numerical) and treat every entry that does not start with a numeric as categorical (also called nominal data specifically). For those starting with a numeric but containing non-numeric character, std::invalid_argument will be thrown.

Definition at line 358 of file CosanData.h.

                                                                                                                 {
  
              std::ifstream indata;
              indata.open(path);
              std::string line;
              std::vector<NumericType> values;
              //    std::vector<std::string> svalues;
              gsl::index rows = 0,cols = 0,col_idx=0;
              //    uint col_idx=0;
              //    std::vector<std::vector<uint>> Idxpinf,Idxminf,Idxmissing;
              //    std::set<uint> colCat;
              NumericType result;
              std::size_t pos;
              std::string SummaryMessage;
  
              //   stod -> "-23","-12E1","-+nan" (is a double type), "+\- inf","+\- infinity", "jklsgfd","1235lkjfg",
 //                 first row: empty/nonempty: empty-> numerical nan,
  //                 not empty-> can read numeric ->  it is indeed a correct numeric format
 //                                                  ->  wrong format i.e. "1235lkjfg" throw error
  //                          -> cannot read numeric->  then consider this as category
  
 //              then set colCat, cols,
              //   numeric-format,
      //
                 std::getline(indata, line);
                 std::stringstream lineStream(line);
                 std::string cell;
                 while(getline(lineStream, cell, ',')) {
                     if (cell.size()==0){
                         values.push_back(StringToNum<NumericType>(std::string("nan")));
                         Idxmissing.push_back(std::vector<gsl::index>({rows,col_idx}));
                         col_idx++;
                         cols=std::max(cols,col_idx);
                         continue;
                     }
                     try{
                         result = StringToNum<NumericType>(cell, &pos);
                     }catch(...){
                         svalues.push_back(cell);
                         colCat.insert(col_idx);
                         col_idx++;
                         cols=std::max(cols,col_idx);
                         continue;
                     }
                     if (pos!=cell.size()){
                         throw std::invalid_argument(
                                 "Incorrect numeric format! Abort the program. The entry reads "+cell+
                                 " and the position is ("+ std::to_string(rows)+","+ std::to_string(col_idx)+")");
                     }
                     values.push_back(result);
                     if (isinf(values.back())){
                         if (values.back()==std::numeric_limits<NumericType>::infinity()){
                             Idxpinf.push_back(std::vector<gsl::index>({rows,col_idx}));}
                         else {Idxminf.push_back(std::vector<gsl::index>({rows,col_idx}));}
                     }
                     else if (isnan(values.back())){
                         Idxmissing.push_back(std::vector<gsl::index>({rows,col_idx}));
                     }
                     col_idx++;
                     cols=std::max(cols,col_idx);
                 }
                 rows = 1;
                 col_idx = 0;
  
  
                 while (std::getline(indata, line)) {
                 //        std::stringstream lineStream(line);
                 //        std::string cell;
                     lineStream.str("");
                     lineStream.clear(); // Clear state flags.
                     lineStream<<line;
                     while(getline(lineStream, cell, ',')) {
                         if (cell.size()==0){
                             if (colCat.find(col_idx)==colCat.end()){
                                 values.push_back(StringToNum<NumericType>(std::string("nan")));
                             }
                             else{
                                 svalues.push_back("");
                             }
                             Idxmissing.push_back(std::vector<gsl::index>({rows,col_idx}));
                             col_idx++;
                             continue;
                         }
                         try{
                             result = StringToNum<NumericType>(cell, &pos);
                         }catch(...){
                             if (colCat.find(col_idx)!=colCat.end())
                             {
                                 svalues.push_back(cell);
                                 colCat.insert(col_idx);
                                 col_idx++;
                                 continue;}
                             else{
                                 throw std::invalid_argument(
                                         "Incorrect value type! Should be numeric but non-numeric input. The entry reads "+cell+
                                         " and the position is ("+ std::to_string(rows)+","+ std::to_string(col_idx)+")");
                             }
                         }
                         if (pos!=cell.size()){
                             throw std::invalid_argument(
                                     "Incorrect numeric format! Abort the program. The entry reads "+cell+
                                     " and the position is ("+ std::to_string(rows)+","+ std::to_string(col_idx)+")");
                         }
                         values.push_back(result);
                         if (isinf(values.back())){
                             if (values.back()==std::numeric_limits<NumericType>::infinity()){
                                 Idxpinf.push_back(std::vector<gsl::index>({rows,col_idx}));}
                             else {Idxminf.push_back(std::vector<gsl::index>({rows,col_idx}));}
                         }
                         else if (isnan(values.back())){
                             Idxmissing.push_back(std::vector<gsl::index>({rows,col_idx}));
                         }
                         col_idx++;
                     }
                     if (cols!=col_idx){
                         std::cout<<cols<<" "<<col_idx<<std::endl;
                         throw std::invalid_argument("Not all rows has same number of entry! First row has "+std::to_string(cols)+" columns but row "+std::to_string(rows)+" has "+std::to_string(col_idx)+" columns!" );
                     }
                     ++rows;
                     col_idx=0;
                 }
                 X.resize(rows,values.size()/rows);
                 gsl::index i =0,__cols =  values.size()/rows;
                 for (auto &each :values ){
                     X(i/__cols,i%__cols) = each;
                     i++;
                 }
  
 //                        Eigen::Map<const CosanMatrix<NumericType>>(values.data(), rows, values.size()/rows);
  
                 SummaryMessage+="Number of rows: "+std::to_string(rows)+"\n";
                 SummaryMessage+="Number of columns: "+std::to_string(cols)+"\n";
                 SummaryMessage+="Number of positive infinity values: "+std::to_string(Idxpinf.size())+". They are at " ;
                 for(auto each :Idxpinf){
                     SummaryMessage+="("+std::to_string(each[0])+","+std::to_string(each[1])+") ";
                 }
                 SummaryMessage+="\n";
                 SummaryMessage+="Number of negative infinity values: "+std::to_string(Idxminf.size())+". They are at ";
                 for(auto each :Idxminf){
                     SummaryMessage+="("+std::to_string(each[0])+","+std::to_string(each[1])+")"+" ";
                 }
                 SummaryMessage+="\n";
                 SummaryMessage+="Number of missing values: "+std::to_string(Idxmissing.size())+". They are at ";
                 for(auto each :Idxmissing){
                     SummaryMessage+="("+std::to_string(each[0])+","+std::to_string(each[1])+")"+" ";
                 }
                 SummaryMessage+="\n";
                 //    for (auto fvalue:values) {std::cout<<fvalue<<std::endl;}
                 SummaryMessage+="Columns of categorical values: Column ";
                 for (auto idx:colCat) {
                     SummaryMessage+=std::to_string(idx)+" ";}
                 SummaryMessage+="\n";
                 gsl::index j = 0 ;
                 for (gsl::index i = 0;i<cols;i++){
                     if (colCat.find(i)==colCat.end()){
                         _raw2numIdx[i] = j;
                         j++;}
                 }
                 j = 0 ;
                 for (gsl::index i = 0;i<cols;i++){
                     if(colCat.find(i)!=colCat.end()){
                         _raw2catIdx[i]=j;
                         j++;
                     }
                 }
  
  
                 return {rows,cols,SummaryMessage};
  
  
                 }

◆ ConcatenateData()

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::ConcatenateData ( const CosanMatrix< NumericType > & inputX )

inline

Concatenate X using CosanMatrix<NumericType> input X. Add new columns.

Parameters

[in] const CosanMatrix<NumericType>& inputX

Add new columns.

Definition at line 94 of file CosanData.h.

                                                                      {
              if (GetrowsX()!=inputX.rows()){
                  throw std::invalid_argument(fmt::format("To concatenate, the number of rows from inputX should match with original X. Current nrow of X is {:}", GetrowsX() ));
              }
              for (gsl::index i = 0 ;i<inputX.cols();i++){
                  X.conservativeResize(X.rows(), X.cols()+1);
                  X.col(X.cols() - 1) = inputX.col(i);
              }
          }

◆ GetcatY()

template<Numeric NumericType>

bool Cosan::CosanRawData< NumericType >::GetcatY ( ) const

inline

True if Y is categorical data type. False otherwise.

Returns: bool

Definition at line 248 of file CosanData.h.

248 {return catY;}

◆ GetcolCatX()

template<Numeric NumericType>

std::set<gsl::index> Cosan::CosanRawData< NumericType >::GetcolCatX ( ) const

inline

Get the column index (in the origin X of csv file) where the column is of categorical type.

Returns: std::set<gsl::index>

Definition at line 237 of file CosanData.h.

237 {return colCatX;}

◆ GetcolCatY()

template<Numeric NumericType>

std::set<gsl::index> Cosan::CosanRawData< NumericType >::GetcolCatY ( ) const

inline

Get the column index (in the origin Y of csv file) where the column is of categorical type.

Returns: std::set<gsl::index>

Note: If Y is one dimension, the size of return should be one. Otherwise it is empty.

Definition at line 243 of file CosanData.h.

243 {return colCatY;}

◆ GetcolsX()

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::GetcolsX ( )

inline

Get the number of columns for X.

Returns: gsl::index

Definition at line 270 of file CosanData.h.

                              {
              colsX = X.cols();
              return colsX;
          }

◆ GetcolsY()

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::GetcolsY ( )

inline

Get the number of columns for Y.

Returns: gsl::index

Definition at line 278 of file CosanData.h.

                              {
              colsY = Y.cols();
              return colsY;
          }

◆ GetIdxminfX()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxminfX ( ) const

inline

Get the position of negative infinity in the origin data X.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 212 of file CosanData.h.

212 {return IdxminfX;}

◆ GetIdxminfY()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxminfY ( ) const

inline

Get the position of negative infinity in the origin data Y.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 227 of file CosanData.h.

227 {return IdxminfY;}

◆ GetIdxmissingX()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxmissingX ( ) const

inline

Get the position of missing in the origin data X.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 217 of file CosanData.h.

217 {return IdxmissingX;}

◆ GetIdxmissingY()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxmissingY ( ) const

inline

Get the position of missing in the origin data Y.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 232 of file CosanData.h.

232 {return IdxmissingY;}

◆ GetIdxpinfX()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxpinfX ( ) const

inline

Get the position of positive infinity in the origin data X.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 207 of file CosanData.h.

207 {return IdxpinfX;}

◆ GetIdxpinfY()

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::GetIdxpinfY ( ) const

inline

Get the position of positive infinity in the origin data Y.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 222 of file CosanData.h.

222 {return IdxpinfY;}

◆ GetInput() [1/2]

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::GetInput ( )

inline

Get a copy of CosanMatrix<NumericType> X.

Definition at line 141 of file CosanData.h.

                                                {
              return X;
          }

◆ GetInput() [2/2]

template<Numeric NumericType>

const CosanMatrix<NumericType>& Cosan::CosanRawData< NumericType >::GetInput ( ) const

inline

Get a const reference to const CosanMatrix<NumericType> X.

Definition at line 153 of file CosanData.h.

                                                             {
              return X;
          }

◆ GetMissingNumber()

template<Numeric NumericType>

std::tuple<gsl::index,gsl::index> Cosan::CosanRawData< NumericType >::GetMissingNumber ( )

inline

Get the total number data information.

Returns: std::tuple<# missing of X,#missing of Y>

Definition at line 166 of file CosanData.h.

                                                       {
              return {X.array().isNaN().template cast<NumericType>().sum(),Y.array().isNaN().template cast<NumericType>().sum()};
          }

◆ GetName()

template<Numeric NumericType>

virtual const std::string Cosan::CosanRawData< NumericType >::GetName ( ) const

inlinevirtual

Get the name of the objects.

Returns: std::string

Reimplemented from Cosan::CosanBO.

Reimplemented in Cosan::CosanData< NumericType >.

Definition at line 176 of file CosanData.h.

176 {return "Raw Data Object.";}

◆ GetRawToCatIdx()

template<Numeric NumericType>

std::unordered_map<gsl::index,gsl::index>& Cosan::CosanRawData< NumericType >::GetRawToCatIdx ( )

inline

Raw data column index to categorical data column index.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 202 of file CosanData.h.

202 {return _raw2catIdx;}

◆ GetRawToNumIdx()

template<Numeric NumericType>

std::unordered_map<gsl::index,gsl::index>& Cosan::CosanRawData< NumericType >::GetRawToNumIdx ( )

inline

Raw data column index to numeric data matrix X column index.

Returns: std::unordered_map<gsl::index,gsl::index>

Definition at line 197 of file CosanData.h.

197 {return _raw2numIdx;}

◆ GetrowsX()

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::GetrowsX ( )

inline

Get the number of rows for X.

Returns: gsl::index

Definition at line 254 of file CosanData.h.

                               {
              rowsX=X.rows();
              return rowsX;
          }

◆ GetrowsY()

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::GetrowsY ( )

inline

Get the number of rows for Y.

Returns: gsl::index

Definition at line 262 of file CosanData.h.

                              {
              rowsY=Y.rows();
              return rowsY;
          }

◆ GetSummaryMessageX()

template<Numeric NumericType>

const std::string& Cosan::CosanRawData< NumericType >::GetSummaryMessageX ( ) const

inline

Get the summary message on reading csv file on X.

Returns: std::string

Definition at line 187 of file CosanData.h.

187 {return SummaryMessageX;}

◆ GetSummaryMessageY()

template<Numeric NumericType>

const std::string& Cosan::CosanRawData< NumericType >::GetSummaryMessageY ( ) const

inline

Get the summary message on reading csv file on Y.

Returns: std::string

Definition at line 192 of file CosanData.h.

192 {return SummaryMessageY;}

◆ GetsvaluesX()

template<Numeric NumericType>

std::vector<std::string> Cosan::CosanRawData< NumericType >::GetsvaluesX ( ) const

inline

Get the vector of categorical data from X. order: row first.

Returns: std::vector<std::string>

Note: it is a std::vector of std::string. Strings are stored row-first.

Definition at line 287 of file CosanData.h.

287 {return svaluesX;}

◆ GetsvaluesY()

template<Numeric NumericType>

std::vector<std::string> Cosan::CosanRawData< NumericType >::GetsvaluesY ( ) const

inline

Get the vector of categorical data from Y. order: row first.

Returns: std::vector<std::string>

Note: it is a std::vector of std::string. Strings are stored row-first.

Definition at line 293 of file CosanData.h.

293 {return svaluesY;}

◆ GetTarget() [1/2]

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::GetTarget ( )

inline

Get a copy of CosanMatrix<NumericType> Y.

Definition at line 147 of file CosanData.h.

                                                 {
              return Y;
          }

◆ GetTarget() [2/2]

template<Numeric NumericType>

const CosanMatrix<NumericType>& Cosan::CosanRawData< NumericType >::GetTarget ( ) const

inline

Get a const reference to const CosanMatrix<NumericType> Y.

Definition at line 159 of file CosanData.h.

                                                              {
              return Y;
          }

◆ GetType()

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::GetType ( )

inline

Returns: an empty CosanMatrix<NumericType> data structure

Note: it is used to determined CosanMatrix<NumericType> data type.

Definition at line 298 of file CosanData.h.

298 {return __TYPE;}

◆ SetInput()

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::SetInput ( const std::string & srcX )

inline

Update input X from csv file.

Parameters

[in] srcX path to the csv file of data X.

Definition at line 74 of file CosanData.h.

                                               {
              std::tie(rowsX,colsX,SummaryMessageX) = _load_csv(srcX,X,IdxpinfX,IdxminfX,IdxmissingX,svaluesX,colCatX);
          }

◆ SetTarget()

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::SetTarget ( const std::string & srcY )

inline

Update target Y from csv file.

Parameters

[in] srcY path to the csv file of data Y.

Definition at line 82 of file CosanData.h.

                                                {
              std::tie(rowsY,colsY,SummaryMessageY) = _load_csv(srcY,Y,IdxpinfY,IdxminfY,IdxmissingY,svaluesY,colCatY);
              if (colCatY.size()!=0){
                  catY=true;
              }
          }

◆ UpdateCat() [1/2]

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::UpdateCat ( const std::vector< std::string > & inputX )

inline

Update categorical vector svaluesX using std::vector<std::string> & inputX.

Parameters

[in] const std::vector<std::string> & inputX

Definition at line 125 of file CosanData.h.

                                                              {
              svaluesX = inputX;
          }

◆ UpdateCat() [2/2]

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::UpdateCat	(	const std::vector< std::string > &	inputX,
		const std::vector< std::string > &	inputY
	)

inline

Update categorical vector svaluesX,svaluesY using std::vector<std::string> & inputX,inputY.

Parameters

[in] const std::vector<std::string> & inputX,inputY

Definition at line 133 of file CosanData.h.

                                                                                                   {
              svaluesX = inputX;
              svaluesY = inputY;
          }

◆ UpdateData() [1/2]

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::UpdateData ( const CosanMatrix< NumericType > & inputX )

inline

Update X using CosanMatrix<NumericType> input X.

Parameters

[in] const CosanMatrix<NumericType>& inputX

Definition at line 108 of file CosanData.h.

                                                                {
             X = inputX;
         }

◆ UpdateData() [2/2]

template<Numeric NumericType>

void Cosan::CosanRawData< NumericType >::UpdateData	(	const CosanMatrix< NumericType > &	inputX,
		const CosanMatrix< NumericType > &	inputY
	)

inline

Update X and Y using CosanMatrix<NumericType> inputX,inputY.

Parameters

[in] const CosanMatrix<NumericType>& inputX,const CosanMatrix<NumericType>& inputY

Definition at line 116 of file CosanData.h.

                                                                                                         {
              X = inputX;
              Y = inputY;
          }

Member Data Documentation

◆ __TYPE

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::__TYPE

protected

Definition at line 308 of file CosanData.h.

◆ _raw2catIdx

template<Numeric NumericType>

std::unordered_map<gsl::index,gsl::index> Cosan::CosanRawData< NumericType >::_raw2catIdx

private

Definition at line 339 of file CosanData.h.

◆ _raw2numIdx

template<Numeric NumericType>

std::unordered_map<gsl::index,gsl::index> Cosan::CosanRawData< NumericType >::_raw2numIdx

private

Definition at line 339 of file CosanData.h.

◆ catY

template<Numeric NumericType>

bool Cosan::CosanRawData< NumericType >::catY = false

protected

true mean respone variable Y is categorical data.

Definition at line 325 of file CosanData.h.

◆ colCatX

template<Numeric NumericType>

std::set<gsl::index> Cosan::CosanRawData< NumericType >::colCatX

protected

column idx in the origin data that is categorical data.

Definition at line 321 of file CosanData.h.

◆ colCatY

template<Numeric NumericType>

std::set<gsl::index> Cosan::CosanRawData< NumericType >::colCatY

protected

Definition at line 321 of file CosanData.h.

◆ colsX

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::colsX = 0

protected

Definition at line 329 of file CosanData.h.

◆ colsY

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::colsY = 0

protected

Definition at line 333 of file CosanData.h.

◆ IdxminfX

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxminfX

protected

Definition at line 316 of file CosanData.h.

◆ IdxminfY

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxminfY

protected

Definition at line 317 of file CosanData.h.

◆ IdxmissingX

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxmissingX

protected

Definition at line 316 of file CosanData.h.

◆ IdxmissingY

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxmissingY

protected

Definition at line 317 of file CosanData.h.

◆ IdxpinfX

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxpinfX

protected

position for positive, negative infinity and missing values.

Definition at line 316 of file CosanData.h.

◆ IdxpinfY

template<Numeric NumericType>

std::vector<std::vector<gsl::index> > Cosan::CosanRawData< NumericType >::IdxpinfY

protected

Definition at line 317 of file CosanData.h.

◆ rowsX

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::rowsX = 0

protected

number of rows.

Definition at line 329 of file CosanData.h.

◆ rowsY

template<Numeric NumericType>

gsl::index Cosan::CosanRawData< NumericType >::rowsY = 0

protected

number of columns

Definition at line 333 of file CosanData.h.

◆ SummaryMessageX

template<Numeric NumericType>

std::string Cosan::CosanRawData< NumericType >::SummaryMessageX

protected

Loading message.

Definition at line 312 of file CosanData.h.

◆ SummaryMessageY

template<Numeric NumericType>

std::string Cosan::CosanRawData< NumericType >::SummaryMessageY

protected

Definition at line 312 of file CosanData.h.

◆ svaluesX

template<Numeric NumericType>

std::vector<std::string> Cosan::CosanRawData< NumericType >::svaluesX

protected

Get the vector of categorical data from original data. order: row first.

Definition at line 337 of file CosanData.h.

◆ svaluesY

template<Numeric NumericType>

std::vector<std::string> Cosan::CosanRawData< NumericType >::svaluesY

protected

Definition at line 337 of file CosanData.h.

◆ X

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::X

protected

Numeric data from origin CSV file for X.

Definition at line 303 of file CosanData.h.

◆ Y

template<Numeric NumericType>

CosanMatrix<NumericType> Cosan::CosanRawData< NumericType >::Y

protected

Numeric data from origin CSV file for Y.

Definition at line 307 of file CosanData.h.

The documentation for this class was generated from the following file:

cosan/data/CosanData.h

Public Member Functions

Protected Attributes

Private Member Functions

Private Attributes

Detailed Description

template<Numeric NumericType> class Cosan::CosanRawData< NumericType >

Constructor & Destructor Documentation

◆ CosanRawData() [1/3]

◆ CosanRawData() [2/3]

◆ CosanRawData() [3/3]

Member Function Documentation

◆ _load_csv()

◆ ConcatenateData()

◆ GetcatY()

◆ GetcolCatX()

◆ GetcolCatY()

◆ GetcolsX()

◆ GetcolsY()

◆ GetIdxminfX()

◆ GetIdxminfY()

◆ GetIdxmissingX()

◆ GetIdxmissingY()

◆ GetIdxpinfX()

◆ GetIdxpinfY()

◆ GetInput() [1/2]

◆ GetInput() [2/2]

◆ GetMissingNumber()

◆ GetName()

◆ GetRawToCatIdx()

◆ GetRawToNumIdx()

◆ GetrowsX()

◆ GetrowsY()

◆ GetSummaryMessageX()

◆ GetSummaryMessageY()

◆ GetsvaluesX()

◆ GetsvaluesY()

◆ GetTarget() [1/2]

◆ GetTarget() [2/2]

◆ GetType()

◆ SetInput()

◆ SetTarget()

◆ UpdateCat() [1/2]

◆ UpdateCat() [2/2]

◆ UpdateData() [1/2]

◆ UpdateData() [2/2]

Member Data Documentation

◆ __TYPE

◆ _raw2catIdx

◆ _raw2numIdx

◆ catY

◆ colCatX

◆ colCatY

◆ colsX

◆ colsY

◆ IdxminfX

◆ IdxminfY

◆ IdxmissingX

◆ IdxmissingY

◆ IdxpinfX

◆ IdxpinfY

◆ rowsX

◆ rowsY

◆ SummaryMessageX

◆ SummaryMessageY

◆ svaluesX

◆ svaluesY

◆ X

◆ Y

template<Numeric NumericType>
class Cosan::CosanRawData< NumericType >