IMP::kmeans::KMeans Class Reference

#include <IMP/kmeans/KMeans.h>

Inheritance diagram for IMP::kmeans::KMeans:

Detailed Description

Class that wraps and provides an interface to the K-means library by David Mount (GPL license), downloaded and adapted to IMP from http://www.cs.umd.edu/~mount/Projects/KMeans/

For a simple usage example, see modules/kmeans/examples/kmeans_example.py

Warning: KMeans has not been well tested yet

Use with caution and please report any bugs found.

Warning: KMeans has not been stabilized and is likely to change without notice.

Definition at line 72 of file kmeans/KMeans.h.

Public Member Functions
	KMeans (const std::string &fname_data, int dim, unsigned int max_nPts)
	KMeans ()
void	add_data_pt (const IMP::Floats &p)
void	clear_data ()
	Clears all data in object. More...
void	execute (unsigned int k, KM_ALG_TYPE alg_type=KM_LLOYDS, int stages=100)
IMP::Ints	get_assignments () const
	Returns the cluster assignment of each data point. More...
IMP::Floats	get_center (unsigned int i) const
	Returns the i'th center. More...
const IMP::Floats &	get_data_point (unsigned int i) const
	Returns the i'th point in the dataset. More...
unsigned int	get_number_of_centers () const
unsigned int	get_number_of_data_points () const
IMP::Floats	get_squared_distance_to_centers () const
virtual std::string	get_type_name () const
virtual ::IMP::VersionInfo	get_version_info () const
	Get information about the module and version of the object. More...
void	print_centers (LogLevel ll=PROGRESS) const
	Print the centers (assuming execute() was applied) to log. More...
Public Member Functions inherited from IMP::Object
virtual void	clear_caches ()
CheckLevel	get_check_level () const
LogLevel	get_log_level () const
void	set_check_level (CheckLevel l)
void	set_log_level (LogLevel l)
	Set the logging level used in this object. More...
void	set_was_used (bool tf) const
void	show (std::ostream &out=std::cout) const
const std::string &	get_name () const
void	set_name (std::string name)

Additional Inherited Members
Protected Member Functions inherited from IMP::Object
	Object (std::string name)
	Construct an object with the given name. More...
virtual void	do_destroy ()

Constructor & Destructor Documentation

IMP::kmeans::KMeans::KMeans	(	const std::string &	fname_data,
		int	dim,
		unsigned int	max_nPts
	)

Initialize the KMeans object with data from fname_data, assuming input data of dimension dim

Parameters

[in]	fname_data	Input filename. Input is assumed to be in textual (ascii) whitespace separated format, with a fixed number of columns dim. Each row represents a single data point of fixed dimension dim. For example, a file with three examples of dimension 4 would look as follows: 10.3 0.7 1.3 11.1 2.1 1.5 20.1 0.2 10.1 0.9 2.1 10.9
[in]	dim	Dimension of each data point
[in]	max_nPts	Maximal number of points to be read from file

IMP::kmeans::KMeans::KMeans

(

)

Empty constructor for all default initializations - object data is not considered initialized after this call

Member Function Documentation

void IMP::kmeans::KMeans::add_data_pt

(

const IMP::Floats &

p

)

Add a data point for clustering.

Parameters

[in]

p

point to be added

void IMP::kmeans::KMeans::clear_data

(

)

Clears all data in object.

void IMP::kmeans::KMeans::execute	(	unsigned int	k,
		KM_ALG_TYPE	alg_type = KM_LLOYDS,
		int	stages = 100
	)

Execute a kmeans algorithm variant on the data points stored.

Parameters

[in]	k	number of clusters
[in]	alg_type	The k-means algorithm variant to use

See Also

KM_ALG_TYPE

Parameters

[in]

stages

Number of k-means iterations

IMP::Ints IMP::kmeans::KMeans::get_assignments

(

)

const

Returns the cluster assignment of each data point.

Definition at line 149 of file kmeans/KMeans.h.

IMP::Floats IMP::kmeans::KMeans::get_center

(

unsigned int

i

)

const

Returns the i'th center.

Must be called only following a successful execute() invocation

Parameters

[in]

i

Center number in range (0,...,k-1)

const IMP::Floats& IMP::kmeans::KMeans::get_data_point

(

unsigned int

i

)

const

Returns the i'th point in the dataset.

Parameters

[in]

i

Center number in range (0,...,nPts-1)

unsigned int IMP::kmeans::KMeans::get_number_of_centers

(

)

const

Returns

The number of centers after a successful execution

Definition at line 164 of file kmeans/KMeans.h.

unsigned int IMP::kmeans::KMeans::get_number_of_data_points

(

)

const

Returns

The number of data points

Definition at line 133 of file kmeans/KMeans.h.

IMP::Floats IMP::kmeans::KMeans::get_squared_distance_to_centers

(

)

const

Returns the squared distance of each data point to its respective cluster center

Definition at line 157 of file kmeans/KMeans.h.

virtual ::IMP::VersionInfo IMP::kmeans::KMeans::get_version_info ( ) const

virtual

Get information about the module and version of the object.

Reimplemented from IMP::Object.

Definition at line 101 of file kmeans/KMeans.h.

void IMP::kmeans::KMeans::print_centers

(

LogLevel

ll = PROGRESS

)

const

Print the centers (assuming execute() was applied) to log.

Parameters

ll	the log level for printout

---

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

• kmeans/KMeans.h