DatasetInfo Class Reference

Information describing a dataset and its features. More...

Collaboration diagram for DatasetInfo:
[legend]

List of all members.

Public Member Functions

const DatasetFeature & feature (index_t attrib_num) const
 Gets information about a particular feature.
const DatasetFeature & feature (index_t attrib_num) const
 Gets information about a particular feature.
ArrayList< DatasetFeature > & features ()
 Gets a mutable list of all features.
ArrayList< DatasetFeature > & features ()
 Gets a mutable list of all features.
void Init (const char *name_in="dataset")
 Initialize a custom dataset.
void Init (const char *name_in="dataset")
 Initialize a custom dataset.
void InitContinuous (index_t n_features, const char *name_in="dataset")
 Initialize an all-continuous dataset;.
void InitContinuous (index_t n_features, const char *name_in="dataset")
 Initialize an all-continuous dataset;.
success_t InitFromArff (TextLineReader *reader, const char *filename="dataset")
 Initialize explicitly from an ARFF file.
success_t InitFromArff (TextLineReader *reader, const char *filename="dataset")
 Initialize explicitly from an ARFF file.
success_t InitFromCsv (TextLineReader *reader, const char *filename="dataset")
 Initialize from a CSV-like file with numbers only, inferring automatically that if the first row has non-numeric characters, it is a header.
success_t InitFromCsv (TextLineReader *reader, const char *filename="dataset")
 Initialize from a CSV-like file with numbers only, inferring automatically that if the first row has non-numeric characters, it is a header.
success_t InitFromFile (TextLineReader *reader, const char *filename="dataset")
 Initializes the header from a file, either CSV or ARFF.
success_t InitFromFile (TextLineReader *reader, const char *filename="dataset")
 Initializes the header from a file, either CSV or ARFF.
bool is_all_continuous () const
 Checks if all parameters are continuous.
bool is_all_continuous () const
 Checks if all parameters are continuous.
index_t n_features () const
 Gets the number of features.
index_t n_features () const
 Gets the number of features.
const char * name () const
 Gets the title of the data set.
const char * name () const
 Gets the title of the data set.
success_t ReadMatrix (TextLineReader *reader, Matrix *matrix) const
 Populates a matrix from a file, given the internal data model.
success_t ReadMatrix (TextLineReader *reader, Matrix *matrix) const
 Populates a matrix from a file, given the internal data model.
success_t ReadPoint (TextLineReader *reader, double *point, bool *is_done) const
 Reads a single vector.
success_t ReadPoint (TextLineReader *reader, double *point, bool *is_done) const
 Reads a single vector.
void set_name (const char *name_in)
 Sets the title of the data set.
void set_name (const char *name_in)
 Sets the title of the data set.
void WriteArffHeader (TextWriter *writer) const
 Writes the header for an ARFF file.
void WriteArffHeader (TextWriter *writer) const
 Writes the header for an ARFF file.
void WriteCsvHeader (const char *sep, TextWriter *writer) const
 Writes header for CSV file.
void WriteCsvHeader (const char *sep, TextWriter *writer) const
 Writes header for CSV file.
void WriteMatrix (const Matrix &matrix, const char *sep, TextWriter *writer) const
 Writes the contents of a matrix to a file.
void WriteMatrix (const Matrix &matrix, const char *sep, TextWriter *writer) const
 Writes the contents of a matrix to a file.

Detailed Description

Information describing a dataset and its features.

Definition at line 219 of file dataset.h.

Member Function Documentation

const DatasetFeature& DatasetInfo::feature ( index_t  attrib_num  )  const [inline]

Gets information about a particular feature.

Definition at line 236 of file dataset.h.

const DatasetFeature& DatasetInfo::feature ( index_t  attrib_num  )  const [inline]

Gets information about a particular feature.

Definition at line 236 of file dataset.h.

Referenced by Dataset::Format(), SimpleCrossValidator< TClassifier >::Init(), InitFromArff(), and WriteArffHeader().

ArrayList<DatasetFeature>& DatasetInfo::features (  )  [inline]

Gets a mutable list of all features.

Definition at line 231 of file dataset.h.

ArrayList<DatasetFeature>& DatasetInfo::features (  )  [inline]

Gets a mutable list of all features.

Definition at line 231 of file dataset.h.

void DatasetInfo::Init ( const char *  name_in = "dataset"  ) 

Initialize a custom dataset.

This assumes you will eventually use the features() to add features.

Parameters:
name_in the dataset title
void DatasetInfo::Init ( const char *  name_in = "dataset"  ) 

Initialize a custom dataset.

This assumes you will eventually use the features() to add features.

Parameters:
name_in the dataset title

Definition at line 127 of file dataset.cc.

References String::Copy(), and ArrayList< TElem >::Init().

Referenced by InitFromArff(), InitFromCsv(), Dataset::InitFromFile(), and InitFromFile().

void DatasetInfo::InitContinuous ( index_t  n_features,
const char *  name_in = "dataset" 
)

Initialize an all-continuous dataset;.

Parameters:
n_features the number of continuous features
name_in the dataset title
void DatasetInfo::InitContinuous ( index_t  n_features,
const char *  name_in = "dataset" 
)

Initialize an all-continuous dataset;.

Parameters:
n_features the number of continuous features
name_in the dataset title

Definition at line 114 of file dataset.cc.

References String::Copy(), ArrayList< TElem >::Init(), and String::InitSprintf().

Referenced by Dataset::AliasMatrix(), Dataset::CopyMatrix(), and Dataset::OwnMatrix().

success_t DatasetInfo::InitFromArff ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initialize explicitly from an ARFF file.

ARFF LIMITATIONS: Values cannot have spaces or commas, even with quotes; 'string' data type not supported (nominal is supported).

You might just use InitFromFile, which will guess the type for you.

This will read only the header information and leave the reader at the first line of data.

success_t DatasetInfo::InitFromArff ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initialize explicitly from an ARFF file.

ARFF LIMITATIONS: Values cannot have spaces or commas, even with quotes; 'string' data type not supported (nominal is supported).

You might just use InitFromFile, which will guess the type for you.

This will read only the header information and leave the reader at the first line of data.

Definition at line 161 of file dataset.cc.

References String::EqualsNoCase(), TextLineReader::Error(), feature(), TextLineReader::Gobble(), ArrayList< TElem >::Init(), Init(), DatasetFeature::InitNominal(), TextLineReader::Peek(), ArrayList< TElem >::PushBack(), set_name(), ArrayList< TElem >::size(), String::Split(), and DatasetFeature::value_names().

Referenced by InitFromFile().

success_t DatasetInfo::InitFromCsv ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initialize from a CSV-like file with numbers only, inferring automatically that if the first row has non-numeric characters, it is a header.

InitFromFile will automatically detect this.

success_t DatasetInfo::InitFromCsv ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initialize from a CSV-like file with numbers only, inferring automatically that if the first row has non-numeric characters, it is a header.

InitFromFile will automatically detect this.

Definition at line 233 of file dataset.cc.

References TextLineReader::Error(), TextLineReader::Gobble(), ArrayList< TElem >::Init(), Init(), String::InitSprintf(), name(), TextLineReader::Peek(), ArrayList< TElem >::PushBack(), ArrayList< TElem >::size(), and String::Split().

Referenced by InitFromFile().

success_t DatasetInfo::InitFromFile ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initializes the header from a file, either CSV or ARFF.

All header lines will be gobbled, so the reader's position will be left at the first line of actual data. You can then read the data with matrix.

success_t DatasetInfo::InitFromFile ( TextLineReader *  reader,
const char *  filename = "dataset" 
)

Initializes the header from a file, either CSV or ARFF.

All header lines will be gobbled, so the reader's position will be left at the first line of actual data. You can then read the data with matrix.

Definition at line 279 of file dataset.cc.

References String::begin(), TextLineReader::Error(), Init(), InitFromArff(), InitFromCsv(), and TextLineReader::Peek().

Referenced by Dataset::InitFromFile().

bool DatasetInfo::is_all_continuous (  )  const

Checks if all parameters are continuous.

bool DatasetInfo::is_all_continuous (  )  const

Checks if all parameters are continuous.

Definition at line 391 of file dataset.cc.

References DatasetFeature::CONTINUOUS, and ArrayList< TElem >::size().

index_t DatasetInfo::n_features (  )  const [inline]

Gets the number of features.

Definition at line 241 of file dataset.h.

References ArrayList< TElem >::size().

index_t DatasetInfo::n_features (  )  const [inline]

Gets the number of features.

Definition at line 241 of file dataset.h.

References ArrayList< TElem >::size().

Referenced by ReadMatrix(), and ReadPoint().

const char* DatasetInfo::name (  )  const [inline]

Gets the title of the data set.

Definition at line 246 of file dataset.h.

const char* DatasetInfo::name (  )  const [inline]

Gets the title of the data set.

Definition at line 246 of file dataset.h.

Referenced by InitFromCsv(), and WriteCsvHeader().

success_t DatasetInfo::ReadMatrix ( TextLineReader *  reader,
Matrix *  matrix 
) const

Populates a matrix from a file, given the internal data model.

ARFF LIMITATIONS: Values cannot have spaces or commas, even with quotes; 'string' data type not supported (nominal is supported).

Parameters:
reader the reader to get lines from
matrix the matrix to store text into
success_t DatasetInfo::ReadMatrix ( TextLineReader *  reader,
Matrix *  matrix 
) const

Populates a matrix from a file, given the internal data model.

ARFF LIMITATIONS: Values cannot have spaces or commas, even with quotes; 'string' data type not supported (nominal is supported).

Parameters:
reader the reader to get lines from
matrix the matrix to store text into

Definition at line 401 of file dataset.cc.

References ArrayList< TElem >::Init(), n_features(), ArrayList< TElem >::PushBackRaw(), ReadPoint(), ArrayList< TElem >::ReleasePtr(), ArrayList< TElem >::Resize(), ArrayList< TElem >::size(), and ArrayList< TElem >::Trim().

Referenced by Dataset::InitFromFile().

success_t DatasetInfo::ReadPoint ( TextLineReader *  reader,
double *  point,
bool *  is_done 
) const

Reads a single vector.

Parameters:
reader the line reader being used
point an array of length n_features()
is_done set to true if we have finished reading the file successfully -- the value of is_done is undefined if the function returns failure!
Returns:
whether reading the line was successful
success_t DatasetInfo::ReadPoint ( TextLineReader *  reader,
double *  point,
bool *  is_done 
) const

Reads a single vector.

Parameters:
reader the line reader being used
point an array of length n_features()
is_done set to true if we have finished reading the file successfully -- the value of is_done is undefined if the function returns failure!
Returns:
whether reading the line was successful

Definition at line 431 of file dataset.cc.

References String::begin(), String::c_str(), String::Copy(), String::end(), TextLineReader::Error(), TextLineReader::Gobble(), TextLineReader::MoreLines(), n_features(), and TextLineReader::Peek().

Referenced by ReadMatrix().

void DatasetInfo::set_name ( const char *  name_in  )  [inline]

Sets the title of the data set.

Definition at line 251 of file dataset.h.

References String::Copy(), and String::Destruct().

void DatasetInfo::set_name ( const char *  name_in  )  [inline]

Sets the title of the data set.

Definition at line 251 of file dataset.h.

References String::Copy(), and String::Destruct().

Referenced by InitFromArff().

void DatasetInfo::WriteArffHeader ( TextWriter *  writer  )  const

Writes the header for an ARFF file.

void DatasetInfo::WriteArffHeader ( TextWriter *  writer  )  const

Writes the header for an ARFF file.

Definition at line 526 of file dataset.cc.

References String::c_str(), feature(), DatasetFeature::n_values(), DatasetFeature::name(), DatasetFeature::NOMINAL, ArrayList< TElem >::size(), DatasetFeature::type(), and DatasetFeature::value_name().

Referenced by Dataset::WriteArff().

void DatasetInfo::WriteCsvHeader ( const char *  sep,
TextWriter *  writer 
) const

Writes header for CSV file.

Parameters:
sep the value separator (use ",\t" for CSV)
writer the text writer to write the header line to
void DatasetInfo::WriteCsvHeader ( const char *  sep,
TextWriter *  writer 
) const

Writes header for CSV file.

Parameters:
sep the value separator (use ",\t" for CSV)
writer the text writer to write the header line to

Definition at line 549 of file dataset.cc.

References name(), and ArrayList< TElem >::size().

Referenced by Dataset::WriteCsv().

void DatasetInfo::WriteMatrix ( const Matrix &  matrix,
const char *  sep,
TextWriter *  writer 
) const

Writes the contents of a matrix to a file.

Parameters:
matrix the matrix
sep the separator (use ",\t" for CSV)
writer the writer to write to
void DatasetInfo::WriteMatrix ( const Matrix &  matrix,
const char *  sep,
TextWriter *  writer 
) const

Writes the contents of a matrix to a file.

Parameters:
matrix the matrix
sep the separator (use ",\t" for CSV)
writer the writer to write to

Definition at line 559 of file dataset.cc.

References ArrayList< TElem >::size().

Referenced by Dataset::WriteArff(), and Dataset::WriteCsv().

The documentation for this class was generated from the following files:
Generated on Mon Jan 24 12:04:40 2011 for FASTlib by  doxygen 1.6.3