Manages Training and Test Data for building Statistical Models and provides functionality for Crossvalidation. Manages data together with surrogate information. The surrogate variables are provided through a vector (see DataManager), and can contain both continuous or categorical data. The surrogate data is provided through files. One file for each dataset, and one file describing the types of surrogates. This file is also an ascii file with space or EOL separated values. Those values are either 0 or 1, standing for respectively categorical or continuous variable. This class does not support any missing data, so each dataset must come with a surrogate data file, all of which must contain the same number of entries as the type-file.
More...
#include <DataManagerWithSurrogates.h>
|
typedef Representer< T > | RepresenterType |
|
typedef
RepresenterType::DatasetPointerType | DatasetPointerType |
|
typedef
RepresenterType::DatasetConstPointerType | DatasetConstPointerType |
|
typedef DataItemWithSurrogates< T > | DataItemWithSurrogatesType |
|
typedef
DataItemWithSurrogatesType::SurrogateTypeVectorType | SurrogateTypeVectorType |
|
typedef Representer< T > | RepresenterType |
|
typedef
RepresenterType::DatasetPointerType | DatasetPointerType |
|
typedef
RepresenterType::DatasetConstPointerType | DatasetConstPointerType |
|
typedef DataItem< T > | DataItemType |
|
typedef DataItemWithSurrogates< T > | DataItemWithSurrogatesType |
|
typedef std::list< const
DataItemType * > | DataItemListType |
|
typedef CrossValidationFold< T > | CrossValidationFoldType |
|
typedef std::list
< CrossValidationFoldType > | CrossValidationFoldListType |
|
|
SurrogateTypeInfoType | m_typeInfo |
|
RepresenterType * | m_representer |
|
DataItemListType | m_DataItemList |
|
template<typename T>
class statismo::DataManagerWithSurrogates< T >
Manages Training and Test Data for building Statistical Models and provides functionality for Crossvalidation. Manages data together with surrogate information. The surrogate variables are provided through a vector (see DataManager), and can contain both continuous or categorical data. The surrogate data is provided through files. One file for each dataset, and one file describing the types of surrogates. This file is also an ascii file with space or EOL separated values. Those values are either 0 or 1, standing for respectively categorical or continuous variable. This class does not support any missing data, so each dataset must come with a surrogate data file, all of which must contain the same number of entries as the type-file.
- See Also
- DataManager
Add a dataset, together with surrogate information
- Parameters
-
datasetFilename | |
datasetURI | (An URI for the dataset. This info is only added to the metadata). |
surrogateFilename | |
Factory method that creates a new instance of a DataManager class
Returns the source filename defining the surrogate types
Get a structure containing the type info: vector of types, and source filename
Get a vector indicating the types of surrogates variables (Categorical vs Continuous)
Loads the information concerning the types of the surrogates variables (categorical=0, continuous=1) => it is assumed to be in a text file with the entries separated by spaces or EOL character
The documentation for this class was generated from the following files: