data-analysis.cpp File Reference

Build histogram file of data, to be plotted with gnuplot. More...

#include "app_common.hpp"
#include "slifis.hpp"
#include "gnuplot_file.hpp"

Include dependency graph for data-analysis.cpp:

Defines
#define	STDERR_FILE_NAME   "stderr_data-analysis.txt"
Functions
void	process_numeric (const DATA_SET &dataset, size_t NbIntervals, size_t idx, vector< size_t > &v_count, vector< double > &v_labels)
	Fills vector v_count with the number of values in field idx of dataset that are in bins between min and max.
int	main (int argc, char **argv)
	see file data-analysis.cpp
void	usage ()
	Prints application usage (arguments)
Variables
const char *	g_plotfile_name = "da.plt"

---

Detailed Description

Build histogram file of data, to be plotted with gnuplot.

• Arguments:
  1. input datafile (.cvs or .arff style)
  2. nb of intervals to consider (10, for instance)
• Description: This app loads the whole datafile, checks for min/max attributes, builds a histogram of 'n' bins (second argument) of the numerical values of datafile. It saves the histogram bins in ".dat" files (as many as there are columns of data). It also generates a gnuplot command-file, that can be used to generates plots of these histograms.

It is useful for quickly showing range and distribution of an unknown data set, so one can build a FIS adapted to the data.

---

Define Documentation

#define STDERR_FILE_NAME   "stderr_data-analysis.txt"

Referenced by main().

---

Function Documentation

void process_numeric	(	const DATA_SET &	dataset,
		size_t	NbIntervals,
		size_t	idx,
		vector< size_t > &	v_count,
		vector< double > &	v_labels
	)

Fills vector v_count with the number of values in field idx of dataset that are in bins between min and max.

v_labels is filled with the bottom numerical value of the bin

References __IN__, __OUT__, slifis::DATA_SET::GetDataPoint(), slifis::DATASET_PROPERTIES::GetMaxValue(), slifis::DATASET_PROPERTIES::GetMinValue(), slifis::DATA_SET::GetNbPoints(), slifis::DATA_SET::GetProperties(), slifis::DATA_POINT::GetValue_f(), and SLIFIS_EPSILON_REAL.

Referenced by main().

int main	(	int	argc,
		char **	argv
	)

see file data-analysis.cpp

References APP_FATAL_ERROR, APP_MSG, APP_NAME, APP_REGULAR_END, slifis::DT_NUMERIC, slifis::DT_STRING_INDEX, g_plotfile_name, slifis::DATAFILE_INFO::GetAttribName(), slifis::DATA_SET::GetFieldType(), slifis::DATAFILE_INFO::GetFileInfo(), slifis::DATA_SET::GetNbClasses(), slifis::DATA_SET::GetNbFields(), slifis::DATA_SET::GetNbPoints(), slifis::DATA_SET::GetProperties(), slifis::DATA_SET::GetStringCount(), slifis::DATA_SET::GetStringValue(), slifis::DATAFILE_INFO::HasAttribNames(), LOG_USER_ERROR, slifis::DATASET_PROPERTIES::Print(), slifis::DATAFILE_INFO::Print(), process_numeric(), slifis::DATA_SET::ReadData(), REOPEN_STDERR, SLIFIS_STAMP_S, STDERR_FILE_NAME, and usage().

void usage

(

)

Prints application usage (arguments)

---

Variable Documentation

const char* g_plotfile_name = "da.plt"

Referenced by main().