Class to encode and decode data encoded with MSNumpress. More...

#include <OpenMS/FORMAT/MSNumpressCoder.h>

Collaboration diagram for MSNumpressCoder:

Classes
struct	NumpressConfig
	Configuration class for MSNumpress. More...

Public Types
enum	NumpressCompression { NONE, LINEAR, PIC, SLOF, SIZE_OF_NUMPRESSCOMPRESSION }
	Names of compression schemes. More...

Public Member Functions
	MSNumpressCoder ()
	default constructor More...

virtual	~MSNumpressCoder ()
	Destructor. More...

void	encodeNP (const std::vector< double > &in, String &result, bool zlib_compression, const NumpressConfig &config)
	Encodes a vector of floating point numbers into a Base64 string using numpress. More...

void	encodeNP (const std::vector< float > &in, String &result, bool zlib_compression, const NumpressConfig &config)
	encodeNP from a float (convert first to double) More...

void	decodeNP (const String &in, std::vector< double > &out, bool zlib_compression, const NumpressConfig &config)
	Decodes a Base64 string to a vector of floating point numbers using numpress. More...

void	encodeNPRaw (const std::vector< double > &in, String &result, const NumpressConfig &config)
	Encode the data vector "in" to a raw byte array. More...

void	decodeNPRaw (const std::string &in, std::vector< double > &out, const NumpressConfig &config)
	Decode the raw byte array "in" to the result vector "out". More...

Static Public Attributes
static const std::string	NamesOfNumpressCompression [SIZE_OF_NUMPRESSCOMPRESSION]

Private Member Functions
void	decodeNPInternal_ (const unsigned char *in, size_t in_size, std::vector< double > &out, const NumpressConfig &config)

Detailed Description

Class to encode and decode data encoded with MSNumpress.

MSNumpress supports three encoding schemata:

Linear (MS:1002312, MS-Numpress linear prediction compression)
Pic (MS:1002313, MS-Numpress positive integer compression)
Slof (MS:1002314, MS-Numpress short logged float compression)

Note that the linear compression scheme only makes sense for monotonically increasing data (such as retention time and m/z) that is often equally spaced. Pic compression only makes sense for positive integers as all data will be rounded to the nearest integer. Slof makes sense for all other data (such as non-integer intensity values).

For more information on the compression schemata, see

Teleman J et al, "Numerical compression schemes for proteomics mass spectrometry data." Mol Cell Proteomics. 2014 Jun;13(6):1537-42. doi: 10.1074/mcp.O114.037879.

Member Enumeration Documentation

◆ NumpressCompression

enum NumpressCompression

Names of compression schemes.

Enumerator
NONE	No compression is applied.
LINEAR	Linear (MS:1002312, MS-Numpress linear prediction compression)
PIC	Pic (MS:1002313, MS-Numpress positive integer compression)
SLOF	Slof (MS:1002314, MS-Numpress short logged float compression)
SIZE_OF_NUMPRESSCOMPRESSION

Constructor & Destructor Documentation

◆ MSNumpressCoder()

MSNumpressCoder ( )

inline

default constructor

◆ ~MSNumpressCoder()

virtual ~MSNumpressCoder ( )

inlinevirtual

Destructor.

Member Function Documentation

◆ decodeNP()

void decodeNP	(	const String &	in,
		std::vector< double > &	out,
		bool	zlib_compression,
		const NumpressConfig &	config
	)

Decodes a Base64 string to a vector of floating point numbers using numpress.

This code is obtained from the proteowizard implementation ./pwiz/pwiz/data/msdata/BinaryDataEncoder.cpp (adapted by Hannes Roest).

This function will first decode the input base64 string (with optional zlib decompression after decoding) and then apply numpress decoding to the data.

Parameters

in	The base64 encoded string
out	The resulting vector of doubles
zlib_compression	Whether to apply zlib de-compression before numpress de-compression
config	The numpress configuration defining the compression strategy

Exceptions

throws Exception::ConversionError if the string cannot be converted

◆ decodeNPInternal_()

void decodeNPInternal_	(	const unsigned char *	in,
		size_t	in_size,
		std::vector< double > &	out,
		const NumpressConfig &	config
	)

private

◆ decodeNPRaw()

void decodeNPRaw	(	const std::string &	in,
		std::vector< double > &	out,
		const NumpressConfig &	config
	)

Decode the raw byte array "in" to the result vector "out".

Note: The string in should *only* contain the data and _no_ extra null terminating byte.

This performs the raw numpress decoding on a raw byte array (not Base64 encoded). Therefore the input string is likely *unsafe* to handle and is basically a byte container.

Please use the safe versions above unless you only have the raw byte arrays.

Parameters

in	The base64 encoded string
out	The resulting vector of doubles
config	The numpress configuration defining the compression strategy

Exceptions

throws Exception::ConversionError if the string cannot be converted

◆ encodeNP() [1/2]

void encodeNP	(	const std::vector< double > &	in,
		String &	result,
		bool	zlib_compression,
		const NumpressConfig &	config
	)

Encodes a vector of floating point numbers into a Base64 string using numpress.

This code is obtained from the proteowizard implementation ./pwiz/pwiz/data/msdata/BinaryDataEncoder.cpp (adapted by Hannes Roest).

This function will first apply the numpress encoding to the data, then encode the result in base64 (with optional zlib compression before base64 encoding).

Note: In case of error, result string is empty

Parameters

in	The vector of floating point numbers to be encoded
result	The resulting string
zlib_compression	Whether to apply zlib compression after numpress compression
config	The numpress configuration defining the compression strategy

◆ encodeNP() [2/2]

void encodeNP	(	const std::vector< float > &	in,
		String &	result,
		bool	zlib_compression,
		const NumpressConfig &	config
	)

encodeNP from a float (convert first to double)

◆ encodeNPRaw()

void encodeNPRaw	(	const std::vector< double > &	in,
		String &	result,
		const NumpressConfig &	config
	)

Encode the data vector "in" to a raw byte array.

Note: In case of error, "result" is given back unmodified; The result is not a string but a raw byte array and may contain zero bytes

This performs the raw numpress encoding on a set of data and does no Base64 encoding on the result. Therefore the result string is likely *unsafe* to handle and is a raw byte array.

Please use the safe versions above unless you need access to the raw byte arrays.

Parameters

in	The vector of floating point numbers to be encoded
result	The resulting string
config	The numpress configuration defining the compression strategy

Member Data Documentation

◆ NamesOfNumpressCompression

const std::string NamesOfNumpressCompression[SIZE_OF_NUMPRESSCOMPRESSION]

static

Classes

Public Types

Public Member Functions

Static Public Attributes

Private Member Functions

Detailed Description

Member Enumeration Documentation

◆ NumpressCompression

Constructor & Destructor Documentation

◆ MSNumpressCoder()

◆ ~MSNumpressCoder()

Member Function Documentation

◆ decodeNP()

◆ decodeNPInternal_()

◆ decodeNPRaw()

◆ encodeNP() [1/2]

◆ encodeNP() [2/2]

◆ encodeNPRaw()

Member Data Documentation

◆ NamesOfNumpressCompression