Laboratoire ICAR » Plateforme TXM

		if (row == null) return false;

		ArrayList<HashMap<String, String>> data = new ArrayList<>();

		ArrayList<String> cols = new ArrayList<>();

			HashMap<String, String> dataline = new HashMap<>();

		ArrayList<ArrayList<String>> data = new ArrayList<>();

// STANDARD DECLARATIONS

package org.txm.macro

import org.kohsuke.args4j.*

import groovy.transform.Field

import org.txm.rcpapplication.swt.widget.parameters.*

import org.txm.searchengine.cqp.clientExceptions.*

import org.txm.searchengine.cqp.corpus.*

import org.txm.searchengine.cqp.corpus.query.*

import org.apache.commons.lang.time.StopWatch

import java.util.Arrays

import org.jfree.chart.renderer.xy.*

import org.jfree.chart.renderer.*

import org.jfree.chart.plot.*

import org.jfree.data.xy.*

import org.jfree.chart.axis.*

import java.awt.*;

import java.awt.geom.*;

import org.jfree.chart.labels.*

import org.txm.ca.core.chartsengine.jfreechart.themes.highcharts.renderers.*

import org.txm.ca.rcp.editors.*

import org.txm.libs.office.ReadODS

import org.txm.ca.core.chartsengine.jfreechart.datasets.*

import org.jfree.chart.renderer.AbstractRenderer

import org.apache.commons.math3.stat.descriptive.*

println "editor: "+editor

if (!(editor instanceof CAEditor)) {

	println "editor is not a CA editor: $editor, Run the macro with F12 when the editor is selected :-)"

	return

}

def chartEditor = editor.getEditors()[0]

def chartComposite = chartEditor.getComposite()

// BEGINNING OF PARAMETERS

// Declare each parameter here

// (available widget types: Query, File, Folder, String, Text, Boolean, Integer, Float and Date)

//@Field @Option(name="query", usage="an example query", widget="Query", required=true, def='[pos="V.*"]')

//def query

//@Field @Option(name="file", usage="an example file", widget="File", required=true, def="C:/Temp/foo.txt")

//def file

//@Field @Option(name="folder", usage="an example folder", widget="Folder", required=false, def="C:/Temp")

//def folder

//@Field @Option(name="date", usage="an example date", widget="Date", required=false, def="1984-09-01")

//def date

//@Field @Option(name="integer", usage="an example integer", widget="Integer", required=false, def="42")

//def integer

@Field @Option(name="QMin", usage="min Q", widget="Float", required=true, def="0.01")

def QMin

@Field @Option(name="CTRXMin", usage="min CTR X", widget="Float", required=true, def="0.001")

def CTRXMin

@Field @Option(name="CTRYMin", usage="min CTR Y", widget="Float", required=true, def="0.001")

def CTRYMin

@Field @Option(name="stackX", usage="stack X", widget="Float", required=true, def="0.0")

def stackX

@Field @Option(name="stackY", usage="stack Y", widget="Float", required=true, def="0.0")

def stackY

//@Field @Option(name="string", usage="an example string", widget="String", required=false, def="hello world!")

//def string

@Field @Option(name="regexFilter", usage="row property form regex", widget="Text", required=false, def="")

def regexFilter

@Field @Option(name="debug", usage="debug (verbose) mode", widget="Boolean", required=true, def="true")

def debug

// La fenêtre de résultats

CAresultWindow = editor

// Le tableau de données d'aide à l'interprétation

// Les données de l'AFC sont manipulables par les méthodes documentées ici http://txm.sourceforge.net/javadoc/TXM/TBX/org/txm/stat/engine/r/function/CA.html

ica = CAresultWindow.getCA()

ca = ica.getCA()

rowNames = ica.getRowNames()

rowCos2 = ca.getRowCos2()

rowContrib = ca.getRowContrib()

colNames = ica.getColNames()

colCos2 = ca.getColCos2()

colContrib = ca.getColContrib()

F1 = ica.getFirstDimension()-1

F2 = ica.getSecondDimension()-1

println sprintf("F1 = %d, F2 = %d, %d initial rows, %d initial cols", F1+1, F2+1, rowNames.length, colNames.length)

def stats = { vector, factor ->

	s = new DescriptiveStatistics()

	(vector.length).times {

		s.addValue(vector[it][factor]);

	}

	println "min       quartile  median    third quartile  max"

	println sprintf("%8f  %8f  %8f  %8f        %8f", s.getMin(), s.getPercentile(25), s.getPercentile(50), s.getPercentile(75), s.getMax())

}

// rows

println "ROWS -----"

println "CTR"+(F1+1)

stats(rowContrib, F1)

println "CTR"+(F2+1)

stats(rowContrib, F2)

println "Cos² "+(F1+1)

stats(rowCos2, F1)

println "Cos² "+(F2+1)

stats(rowCos2, F2)

// cols

println "COLS -----"

println "CTR"+(F1+1)

stats(colContrib, F1)

println "CTR"+(F2+1)

stats(colContrib, F2)

println "Cos² "+(F1+1)

stats(colCos2, F1)

println "Cos² "+(F2+1)

stats(colCos2, F2)

def stats2 = { vector, factor1, factor2 ->

	s = new DescriptiveStatistics()

	(vector.length).times {

		s.addValue(vector[it][factor1]+vector[it][factor2]);

	}

	println "min       quartile  median    third quartile  max"

	println sprintf("%8f  %8f  %8f  %8f        %8f", s.getMin(), s.getPercentile(25), s.getPercentile(50), s.getPercentile(75), s.getMax())

}

println "Q"+(F1+1)+(F2+1)

stats2(rowCos2, F1, F2)

println "Q"+(F1+1)+(F2+1)

stats2(colCos2, F1, F2)

// Open the parameters input dialog box

if (!ParametersDialog.open(this)) return

// END OF PARAMETERS

filteredRows = []

println ""

if (debug) println "\nROWS -----"

if (regexFilter.length() > 0) {

	nmatched = 0

	rowNames.each { it ->

		if (it ==~ regexFilter) {

			if (debug && !filteredRows.contains(it)) {

				if (nmatched > 0) print ", "

				print it

			}

			filteredRows << it

			nmatched++

		}

	}

	if (debug && nmatched > 0) println ""

	if (debug) {

		println sprintf("\n%d rows filtered by regex", nmatched)

	}

}

nQCTR = 0

rowNames.eachWithIndex { it, i ->

	def cos2 = rowCos2[i]

	if ((rowContrib[i][F1] < CTRXMin) && (rowContrib[i][F2] < CTRYMin) && (cos2[F1] + cos2[F2] < QMin)) {

		if (debug && !filteredRows.contains(rowNames[i])) {

			if (nQCTR > 0) print ", "

			print rowNames[i]

		}

		filteredRows << rowNames[i]

		nQCTR++

	}

}

if (debug) {

	println sprintf("\n%d rows filtered by CTR or Q"+(F1+1)+(F2+1), nQCTR)

}

// cols

filteredCols = []

println ""

if (debug) println "\nCOLS -----"

if (regexFilter.length() > 0) {

	nmatched = 0

	colNames.each { it ->

		if (it ==~ regexFilter) {

			if (debug && !filteredCols.contains(it)) {

				if (nmatched > 0) print ", "

				print it

			}

			filteredCols << it

			nmatched++

		}

	}

	if (debug && nmatched > 0) println ""

	if (debug) {

		println sprintf("\n%d cols filtered by regex", nmatched)

	}

}

nQCTR = 0

colNames.eachWithIndex { it, i ->

	def cos2 = colCos2[i]

	if (!((colContrib[i][F1] > CTRXMin) || (colContrib[i][F2] > CTRYMin) || (cos2[F1] + cos2[F2] > QMin))) {

		if (debug && !filteredCols.contains(colNames[i])) {

			if (nQCTR > 0) print ", "

			print colNames[i]

		}

		filteredCols << colNames[i]

		nQCTR++

	}

}

if (debug) {

	println sprintf("\n%d cols filtered by CTR or Q"+(F1+1)+(F2+1), nQCTR)

}

double stackXv = stackX

double stackYv = stackY

// Visualisation graphique

chartCAresultWindow = CAresultWindow.getEditors()[0]

chartComposite = chartCAresultWindow.getComposite()

monitor.syncExec( new Runnable() {

	public void run() {

				println chartComposite

				def chart = chartEditor.getChart();

				dataset2 = new CAXYDataset(ica)

				labels = dataset2.rowLabels

				coords = dataset2.rowCoordinates

				(labels.length).times {

					if (filteredRows.contains(labels[it])) {

						println "Moving "+labels[it]+" row to origin."

						labels[it] = ""

						coords[it][F1] = stackXv

						coords[it][F2] = stackYv

					}

				}

				labels = dataset2.columnLabels

				coords = dataset2.columnCoordinates

				(labels.length).times {

					if (filteredCols.contains(labels[it])) {

						println "Moving "+labels[it]+" col to origin."

						labels[it] = ""

						coords[it][F1] = stackXv

						coords[it][F2] = stackYv

					}

				}

				chart.getXYPlot().setDataset(dataset2)

				ica.getChartCreator().getChartsEngine().getJFCTheme().apply(chart); // need to be call AFTER setRenderer() cause this method changes some renderering parameters

				chartComposite.loadChart()

			}

})