mio::ResamplingAlgorithms Class Reference

Detailed Description

Interface class for the temporal resampling algorithms.

These models generate data points that are missing based on neighbouring points in a time series.

Author

Mathias Bavay - Thomas Egger

Date

2013-05-24

#include <ResamplingAlgorithms.h>

Public Types
enum	ResamplingPosition { hasTimestamp , inGap , begin , end }
using	gap_info = GapInfo

Public Member Functions
	ResamplingAlgorithms (const std::string &i_algoname, const std::string &i_parname, const double &dflt_max_gap_size, const std::vector< std::pair< std::string, std::string > > &)
virtual	~ResamplingAlgorithms ()
gap_info	findGap (const std::string &stationHash, const size_t &pos, const size_t &paramindex, const std::vector< MeteoData > &vecM, const Date &resampling_date, const double &i_max_gap_size)
	Find a data gap around a given position.
virtual bool	resample (const std::string &stationHash, const size_t &index, const ResamplingPosition &position, const size_t &paramindex, const std::vector< MeteoData > &vecM, MeteoData &md, const gap_info &gap, const bool &exactMatch)=0
void	resetResampling ()
const std::string	getAlgo () const
double	getMaxGapSize () const
virtual std::string	toString () const =0

Static Protected Member Functions
static double	partialAccumulateAtLeft (const std::vector< MeteoData > &vecM, const size_t &paramindex, const size_t &pos, const Date &curr_date)
	‍For algorithms that support extrapolations, this could be set to true in the ini file
static double	partialAccumulateAtRight (const std::vector< MeteoData > &vecM, const size_t &paramindex, const size_t &pos, const Date &curr_date)
static double	linearInterpolation (const double &x1, const double &y1, const double &x2, const double &y2, const double &x3)
	This function solves the equation y = ax + b for two given points and returns y for a given x.
static Date	getDailyStart (const Date &resampling_date)
	For a given date, find the start of the day, considering that for midnight we return the day before! (as is necessary for daily averages, sums, etc that can be provided at midnight for the day before)
static size_t	getDailyValue (const std::vector< MeteoData > &vecM, const size_t &paramindex, size_t pos, const Date &intervalStart, const Date &intervalEnd)
	Find a unique value in a given time interval. This is useful for retrieving a unique daily average, daily sum, etc.

Protected Attributes
bool	allowExtrapolate = false
const std::string	algo
const std::string	parname
double	max_gap_size
	Maximum extend of a data gap over which to interpolate.

Member Typedef Documentation

gap_info

using mio::ResamplingAlgorithms::gap_info = GapInfo

Member Enumeration Documentation

ResamplingPosition

enum mio::ResamplingAlgorithms::ResamplingPosition

Enumerator
hasTimestamp
inGap
begin
end

Constructor & Destructor Documentation

ResamplingAlgorithms()

mio::ResamplingAlgorithms::ResamplingAlgorithms ( const std::string &  i_algoname, const std::string &  i_parname, const double &  dflt_max_gap_size, const std::vector< std::pair< std::string, std::string > > &    )

inline

~ResamplingAlgorithms()

virtual mio::ResamplingAlgorithms::~ResamplingAlgorithms ( )

inlinevirtual

Member Function Documentation

findGap()

ResamplingAlgorithms::gap_info mio::ResamplingAlgorithms::findGap	(	const std::string &	stationHash,
		const size_t &	pos,
		const size_t &	paramindex,
		const std::vector< MeteoData > &	vecM,
		const Date &	resampling_date,
		const double &	i_max_gap_size
	)

Find a data gap around a given position.

Parameters

[in]	stationHash	Station hash used to uniquely identify the station (for caching)
[in]	pos	Current position (index at requested date or next point after)
[in]	paramindex	Meteo parameter index
[in]	vecM	Timeseries of MeteoData for the given station
[in]	resampling_date	Date to resample (ie requested date)
[in]	i_max_gap_size	Maximum gap size to search (in Julian days)

Returns

Gap metadata (last point before, first point after, indices, dates, size)

getAlgo()

const std::string mio::ResamplingAlgorithms::getAlgo ( ) const

inline

getDailyStart()

Date mio::ResamplingAlgorithms::getDailyStart ( const Date &  resampling_date )

staticprotected

For a given date, find the start of the day, considering that for midnight we return the day before! (as is necessary for daily averages, sums, etc that can be provided at midnight for the day before)

Parameters

[in]

resampling_date

current date

Returns

start of the day or start of the day before in case of midnight

getDailyValue()

size_t mio::ResamplingAlgorithms::getDailyValue ( const std::vector< MeteoData > &  vecM, const size_t &  paramindex, size_t  pos, const Date &  intervalStart, const Date &  intervalEnd  )

staticprotected

Find a unique value in a given time interval. This is useful for retrieving a unique daily average, daily sum, etc.

Parameters

[in]	vecM	Timeseries of MeteoData for the given station
[in]	paramindex	index of the meteo parameter to process
[in]	pos	index of the current time step
[in]	intervalStart	start of the interval within which we will look for a valid value (often, start of the day)
[in]	intervalEnd	end of the interval within which we will look for a valid value (often, end of the day)

Returns

index of the parameter's valid value

getMaxGapSize()

double mio::ResamplingAlgorithms::getMaxGapSize ( ) const

inline

linearInterpolation()

double mio::ResamplingAlgorithms::linearInterpolation ( const double &  x1, const double &  y1, const double &  x2, const double &  y2, const double &  x  )

staticprotected

This function solves the equation y = ax + b for two given points and returns y for a given x.

Parameters

[in]	x1	x-coordinate of first point
[in]	y1	y-coordinate of first point
[in]	x2	x-coordinate of second point
[in]	y2	y-coordinate of second point
[in]	x	x-coordinate of desired point

Returns

y-coordinate of desired point

partialAccumulateAtLeft()

double mio::ResamplingAlgorithms::partialAccumulateAtLeft ( const std::vector< MeteoData > &  vecM, const size_t &  paramindex, const size_t &  pos, const Date &  curr_date  )

staticprotected

‍For algorithms that support extrapolations, this could be set to true in the ini file

partialAccumulateAtRight()

double mio::ResamplingAlgorithms::partialAccumulateAtRight ( const std::vector< MeteoData > &  vecM, const size_t &  paramindex, const size_t &  pos, const Date &  curr_date  )

staticprotected

resample()

virtual bool mio::ResamplingAlgorithms::resample ( const std::string &  stationHash, const size_t &  index, const ResamplingPosition &  position, const size_t &  paramindex, const std::vector< MeteoData > &  vecM, MeteoData &  md, const gap_info &  gap, const bool &  exactMatch  )

pure virtual

Implemented in mio::Accumulate, mio::ARIMAResampling, mio::DailyAverage, mio::Daily_solar, mio::LinearResampling, mio::NearestNeighbour, mio::NoResampling, mio::RegressionFill, and mio::Solar.

resetResampling()

void mio::ResamplingAlgorithms::resetResampling ( )

inline

toString()

virtual std::string mio::ResamplingAlgorithms::toString ( ) const

pure virtual

Implemented in mio::Accumulate, mio::ARIMAResampling, mio::DailyAverage, mio::Daily_solar, mio::LinearResampling, mio::NearestNeighbour, mio::NoResampling, mio::RegressionFill, and mio::Solar.

Member Data Documentation

algo

const std::string mio::ResamplingAlgorithms::algo

protected

allowExtrapolate

bool mio::ResamplingAlgorithms::allowExtrapolate = false

protected

max_gap_size

double mio::ResamplingAlgorithms::max_gap_size

protected

Maximum extend of a data gap over which to interpolate.

parname

const std::string mio::ResamplingAlgorithms::parname

protected

---

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

• ResamplingAlgorithms.h
• ResamplingAlgorithms.cc