#include "catchdistribution.h"
#include "readfunc.h"
#include "readword.h"
#include "readaggregation.h"
#include "errorhandler.h"
#include "areatime.h"
#include "fleet.h"
#include "stock.h"
#include "multinomial.h"
#include "mathfunc.h"
#include "stockprey.h"
#include "ludecomposition.h"
#include "gadget.h"
#include "global.h"
CatchDistribution::CatchDistribution(CommentStream& infile, const AreaClass* const Area,
const TimeClass* const TimeInfo, Keeper* const keeper, double weight, const char* name)
: Likelihood(CATCHDISTRIBUTIONLIKELIHOOD, weight, name), alptr(0) {
int i, j;
char text[MaxStrLength];
strncpy(text, "", MaxStrLength);
int numarea = 0, numage = 0, numlen = 0;
char datafilename[MaxStrLength];
char aggfilename[MaxStrLength];
strncpy(datafilename, "", MaxStrLength);
strncpy(aggfilename, "", MaxStrLength);
ifstream datafile;
CommentStream subdata(datafile);
timeindex = 0;
yearly = 0;
functionname = new char[MaxStrLength];
strncpy(functionname, "", MaxStrLength);
readWordAndValue(infile, "datafile", datafilename);
readWordAndValue(infile, "function", functionname);
functionnumber = 0;
if (strcasecmp(functionname, "multinomial") == 0) {
MN = Multinomial();
functionnumber = 1;
} else if (strcasecmp(functionname, "pearson") == 0) {
functionnumber = 2;
} else if (strcasecmp(functionname, "gamma") == 0) {
functionnumber = 3;
} else if (strcasecmp(functionname, "sumofsquares") == 0) {
functionnumber = 4;
} else if (strcasecmp(functionname, "mvn") == 0) {
functionnumber = 5;
readWordAndVariable(infile, "lag", lag);
readWordAndVariable(infile, "sigma", sigma);
sigma.Inform(keeper);
params.resize(lag, keeper);
for (i = 0; i < lag; i++)
readWordAndVariable(infile, "param", params[i]);
params.Inform(keeper);
} else if (strcasecmp(functionname, "mvlogistic") == 0) {
functionnumber = 6;
readWordAndVariable(infile, "sigma", sigma);
sigma.Inform(keeper);
} else if (strcasecmp(functionname, "log") == 0) {
//JMB moved the logcatch function to here instead of it being a seperate class
functionnumber = 7;
} else if (strcasecmp(functionname, "stratified") == 0) {
//JMB experimental version of the sum of squares function for stratified samples
functionnumber = 8;
} else
handle.logFileMessage(LOGFAIL, "\nError in catchdistribution - unrecognised function", functionname);
infile >> ws;
char c = infile.peek();
if ((c == 'a') || (c == 'A')) {
//we have found either aggregationlevel or areaaggfile ...
streampos pos = infile.tellg();
infile >> text >> ws;
if ((strcasecmp(text, "aggregation_level") == 0) || (strcasecmp(text, "aggregationlevel") == 0))
infile >> yearly >> ws;
else if (strcasecmp(text, "areaaggfile") == 0)
infile.seekg(pos);
else
handle.logFileUnexpected(LOGFAIL, "areaaggfile", text);
//JMB - peek at the next char
c = infile.peek();
if (yearly != 0 && yearly != 1)
handle.logFileMessage(LOGFAIL, "\nError in catchdistribution - aggregationlevel must be 0 or 1");
}
//JMB - changed to make the reading of overconsumption optional
if ((c == 'o') || (c == 'O')) {
readWordAndVariable(infile, "overconsumption", overconsumption);
infile >> ws;
c = infile.peek();
} else
overconsumption = 0;
if (overconsumption != 0 && overconsumption != 1)
handle.logFileMessage(LOGFAIL, "\nError in catchdistribution - overconsumption must be 0 or 1");
//JMB - changed to make the reading of minimum probability optional
if ((c == 'm') || (c == 'M'))
readWordAndVariable(infile, "minimumprobability", epsilon);
else if ((c == 'e') || (c == 'E'))
readWordAndVariable(infile, "epsilon", epsilon);
else
epsilon = 10.0;
if (epsilon < verysmall) {
handle.logFileMessage(LOGWARN, "epsilon should be a positive integer - set to default value 10");
epsilon = 10.0;
}
//read in area aggregation from file
readWordAndValue(infile, "areaaggfile", aggfilename);
datafile.open(aggfilename, ios::in);
handle.checkIfFailure(datafile, aggfilename);
handle.Open(aggfilename);
numarea = readAggregation(subdata, areas, areaindex);
handle.Close();
datafile.close();
datafile.clear();
//read in age aggregation from file
readWordAndValue(infile, "ageaggfile", aggfilename);
datafile.open(aggfilename, ios::in);
handle.checkIfFailure(datafile, aggfilename);
handle.Open(aggfilename);
numage = readAggregation(subdata, ages, ageindex);
handle.Close();
datafile.close();
datafile.clear();
//read in length aggregation from file
readWordAndValue(infile, "lenaggfile", aggfilename);
datafile.open(aggfilename, ios::in);
handle.checkIfFailure(datafile, aggfilename);
handle.Open(aggfilename);
numlen = readLengthAggregation(subdata, lengths, lenindex);
handle.Close();
datafile.close();
datafile.clear();
LgrpDiv = new LengthGroupDivision(lengths);
if (LgrpDiv->Error())
handle.logMessage(LOGFAIL, "Error in catchdistribution - failed to create length group");
//Must change from outer areas to inner areas.
for (i = 0; i < areas.Nrow(); i++)
for (j = 0; j < areas.Ncol(i); j++)
areas[i][j] = Area->getInnerArea(areas[i][j]);
//read in the fleetnames
i = 0;
infile >> text >> ws;
if (strcasecmp(text, "fleetnames") != 0)
handle.logFileUnexpected(LOGFAIL, "fleetnames", text);
infile >> text >> ws;
while (!infile.eof() && (strcasecmp(text, "stocknames") != 0)) {
fleetnames.resize(new char[strlen(text) + 1]);
strcpy(fleetnames[i++], text);
infile >> text >> ws;
}
if (fleetnames.Size() == 0)
handle.logFileMessage(LOGFAIL, "\nError in catchdistribution - failed to read fleets");
handle.logMessage(LOGMESSAGE, "Read fleet data - number of fleets", fleetnames.Size());
//read in the stocknames
i = 0;
if (strcasecmp(text, "stocknames") != 0)
handle.logFileUnexpected(LOGFAIL, "stocknames", text);
infile >> text;
while (!infile.eof() && (strcasecmp(text, "[component]") != 0)) {
infile >> ws;
stocknames.resize(new char[strlen(text) + 1]);
strcpy(stocknames[i++], text);
infile >> text;
}
if (stocknames.Size() == 0)
handle.logFileMessage(LOGFAIL, "\nError in catchdistribution - failed to read stocks");
handle.logMessage(LOGMESSAGE, "Read stock data - number of stocks", stocknames.Size());
//We have now read in all the data from the main likelihood file
//But we have to read in the statistics data from datafilename
datafile.open(datafilename, ios::in);
handle.checkIfFailure(datafile, datafilename);
handle.Open(datafilename);
readDistributionData(subdata, TimeInfo, numarea, numage, numlen);
handle.Close();
datafile.close();
datafile.clear();
switch (functionnumber) {
case 2:
case 3:
case 4:
case 7:
case 8:
for (i = 0; i < numarea; i++) {
modelYearData.resize(new DoubleMatrix(numage, numlen, 0.0));
obsYearData.resize(new DoubleMatrix(numage, numlen, 0.0));
}
break;
case 1:
case 5:
case 6:
if (yearly)
handle.logMessage(LOGWARN, "Warning in catchdistribution - yearly aggregation is ignored for function", functionname);
yearly = 0;
break;
default:
handle.logMessage(LOGWARN, "Warning in catchdistribution - unrecognised function", functionname);
break;
}
}
void CatchDistribution::readDistributionData(CommentStream& infile,
const TimeClass* TimeInfo, int numarea, int numage, int numlen) {
int i, year, step;
double tmpnumber;
char tmparea[MaxStrLength], tmpage[MaxStrLength], tmplen[MaxStrLength];
strncpy(tmparea, "", MaxStrLength);
strncpy(tmpage, "", MaxStrLength);
strncpy(tmplen, "", MaxStrLength);
int keepdata, timeid, ageid, areaid, lenid, count, reject;
//Check the number of columns in the inputfile
infile >> ws;
if (countColumns(infile) != 6)
handle.logFileMessage(LOGFAIL, "wrong number of columns in inputfile - should be 6");
year = step = count = reject = 0;
while (!infile.eof()) {
keepdata = 1;
infile >> year >> step >> tmparea >> tmpage >> tmplen >> tmpnumber >> ws;
//crude check to see if something has gone wrong and avoid infinite loops
if (strlen(tmparea) == 0)
handle.logFileMessage(LOGFAIL, "failed to read data from file");
//if tmparea is in areaindex find areaid, else dont keep the data
areaid = -1;
for (i = 0; i < areaindex.Size(); i++)
if (strcasecmp(areaindex[i], tmparea) == 0)
areaid = i;
if (areaid == -1)
keepdata = 0;
//if tmpage is in ageindex find ageid, else dont keep the data
ageid = -1;
for (i = 0; i < ageindex.Size(); i++)
if (strcasecmp(ageindex[i], tmpage) == 0)
ageid = i;
if (ageid == -1)
keepdata = 0;
//if tmplen is in lenindex find lenid, else dont keep the data
lenid = -1;
for (i = 0; i < lenindex.Size(); i++)
if (strcasecmp(lenindex[i], tmplen) == 0)
lenid = i;
if (lenid == -1)
keepdata = 0;
//check if the year and step are in the simulation
timeid = -1;
if ((TimeInfo->isWithinPeriod(year, step)) && (keepdata == 1)) {
//if this is a new timestep, resize to store the data
for (i = 0; i < Years.Size(); i++)
if ((Years[i] == year) && (Steps[i] == step))
timeid = i;
if (timeid == -1) {
Years.resize(1, year);
Steps.resize(1, step);
timeid = (Years.Size() - 1);
obsDistribution.resize();
modelDistribution.resize();
likelihoodValues.AddRows(1, numarea, 0.0);
for (i = 0; i < numarea; i++) {
obsDistribution[timeid].resize(new DoubleMatrix(numage, numlen, 0.0));
modelDistribution[timeid].resize(new DoubleMatrix(numage, numlen, 0.0));
}
}
} else
keepdata = 0;
if (keepdata == 1) {
//distribution data is required, so store it
count++;
(*obsDistribution[timeid][areaid])[ageid][lenid] = tmpnumber;
} else
reject++; //count number of rejected data points read from file
}
AAT.addActions(Years, Steps, TimeInfo);
if (count == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - found no data in the data file for", this->getName());
if (reject != 0)
handle.logMessage(LOGMESSAGE, "Discarded invalid catchdistribution data - number of invalid entries", reject);
handle.logMessage(LOGMESSAGE, "Read catchdistribution data file - number of entries", count);
}
CatchDistribution::~CatchDistribution() {
int i, j;
for (i = 0; i < stocknames.Size(); i++)
delete[] stocknames[i];
for (i = 0; i < fleetnames.Size(); i++)
delete[] fleetnames[i];
for (i = 0; i < areaindex.Size(); i++)
delete[] areaindex[i];
for (i = 0; i < ageindex.Size(); i++)
delete[] ageindex[i];
for (i = 0; i < lenindex.Size(); i++)
delete[] lenindex[i];
for (i = 0; i < obsDistribution.Nrow(); i++)
for (j = 0; j < obsDistribution.Ncol(i); j++) {
delete obsDistribution[i][j];
delete modelDistribution[i][j];
}
for (i = 0; i < modelYearData.Size(); i++) {
delete modelYearData[i];
delete obsYearData[i];
}
delete aggregator;
delete LgrpDiv;
delete[] functionname;
}
void CatchDistribution::Reset(const Keeper* const keeper) {
Likelihood::Reset(keeper);
if (isZero(weight))
handle.logMessage(LOGWARN, "Warning in catchdistribution - zero weight for", this->getName());
int i, j;
for (i = 0; i < modelDistribution.Nrow(); i++)
for (j = 0; j < modelDistribution.Ncol(i); j++)
(*modelDistribution[i][j]).setToZero();
if (yearly)
for (i = 0; i < modelYearData.Size(); i++) {
(*modelYearData[i]).setToZero();
(*obsYearData[i]).setToZero();
}
switch (functionnumber) {
case 2:
case 3:
case 4:
case 6:
case 7:
case 8:
break;
case 1:
MN.setValue(epsilon);
break;
case 5:
illegal = 0;
this->calcCorrelation();
if ((illegal) || (LU.isIllegal()))
handle.logMessage(LOGWARN, "Warning in catchdistribution - multivariate normal out of bounds");
break;
default:
handle.logMessage(LOGWARN, "Warning in catchdistribution - unrecognised function", functionname);
break;
}
if (handle.getLogLevel() >= LOGMESSAGE)
handle.logMessage(LOGMESSAGE, "Reset catchdistribution component", this->getName());
}
void CatchDistribution::Print(ofstream& outfile) const {
int i;
outfile << "\nCatch Distribution " << this->getName() << " - likelihood value " << likelihood
<< "\n\tFunction " << functionname << "\n\tStock names:";
for (i = 0; i < stocknames.Size(); i++)
outfile << sep << stocknames[i];
outfile << "\n\tFleet names:";
for (i = 0; i < fleetnames.Size(); i++)
outfile << sep << fleetnames[i];
outfile << endl;
switch (functionnumber) {
case 1:
case 2:
case 3:
case 4:
case 7:
case 8:
break;
case 5:
outfile << "\tMultivariate normal distribution parameters: sigma " << sigma;
for (i = 0; i < lag; i++)
outfile << " param" << i + 1 << " " << params[i];
outfile << endl;
break;
case 6:
outfile << "\tMultivariate logistic distribution parameter: sigma " << sigma << endl;
break;
default:
handle.logMessage(LOGWARN, "Warning in catchdistribution - unrecognised function", functionname);
break;
}
aggregator->Print(outfile);
outfile.flush();
}
void CatchDistribution::printLikelihood(ofstream& outfile, const TimeClass* const TimeInfo) {
if (!AAT.atCurrentTime(TimeInfo))
return;
int i, area, age, len;
timeindex = -1;
for (i = 0; i < Years.Size(); i++)
if ((Years[i] == TimeInfo->getYear()) && (Steps[i] == TimeInfo->getStep()))
timeindex = i;
if (timeindex == -1)
handle.logMessage(LOGFAIL, "Error in catchdistribution - invalid timestep");
for (area = 0; area < modelDistribution.Ncol(timeindex); area++) {
for (age = 0; age < modelDistribution[timeindex][area]->Nrow(); age++) {
for (len = 0; len < modelDistribution[timeindex][area]->Ncol(age); len++) {
outfile << setw(lowwidth) << Years[timeindex] << sep << setw(lowwidth)
<< Steps[timeindex] << sep << setw(printwidth) << areaindex[area] << sep
<< setw(printwidth) << ageindex[age] << sep << setw(printwidth)
<< lenindex[len] << sep << setprecision(largeprecision) << setw(largewidth);
//JMB crude filter to remove the 'silly' values from the output
if ((*modelDistribution[timeindex][area])[age][len] < rathersmall)
outfile << 0 << endl;
else
outfile << (*modelDistribution[timeindex][area])[age][len] << endl;
}
}
}
}
void CatchDistribution::setFleetsAndStocks(FleetPtrVector& Fleets, StockPtrVector& Stocks) {
int i, j, k, found, minage, maxage;
FleetPtrVector fleets;
StockPtrVector stocks;
for (i = 0; i < fleetnames.Size(); i++) {
found = 0;
for (j = 0; j < Fleets.Size(); j++) {
if (strcasecmp(fleetnames[i], Fleets[j]->getName()) == 0) {
found ++;
fleets.resize(Fleets[j]);
}
}
if (found == 0)
handle.logMessage(LOGFAIL, "Error in catchdistribution - unrecognised fleet", fleetnames[i]);
}
for (i = 0; i < fleets.Size(); i++)
for (j = 0; j < fleets.Size(); j++)
if ((strcasecmp(fleets[i]->getName(), fleets[j]->getName()) == 0) && (i != j))
handle.logMessage(LOGFAIL, "Error in catchdistribution - repeated fleet", fleets[i]->getName());
for (i = 0; i < stocknames.Size(); i++) {
found = 0;
for (j = 0; j < Stocks.Size(); j++) {
if (Stocks[j]->isEaten()) {
if (strcasecmp(stocknames[i], Stocks[j]->getName()) == 0) {
found++;
stocks.resize(Stocks[j]);
}
}
}
if (found == 0)
handle.logMessage(LOGFAIL, "Error in catchdistribution - unrecognised stock", stocknames[i]);
}
for (i = 0; i < stocks.Size(); i++)
for (j = 0; j < stocks.Size(); j++)
if ((strcasecmp(stocks[i]->getName(), stocks[j]->getName()) == 0) && (i != j))
handle.logMessage(LOGFAIL, "Error in catchdistribution - repeated stock", stocks[i]->getName());
//check fleet areas and stock areas, ages and lengths
if (handle.getLogLevel() >= LOGWARN) {
for (j = 0; j < areas.Nrow(); j++) {
found = 0;
for (i = 0; i < fleets.Size(); i++)
for (k = 0; k < areas.Ncol(j); k++)
if (fleets[i]->isInArea(areas[j][k]))
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - fleet not defined on all areas");
}
for (j = 0; j < areas.Nrow(); j++) {
found = 0;
for (i = 0; i < stocks.Size(); i++)
for (k = 0; k < areas.Ncol(j); k++)
if (stocks[i]->isInArea(areas[j][k]))
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - stock not defined on all areas");
}
minage = 9999;
maxage = 0;
for (i = 0; i < ages.Nrow(); i++) {
for (j = 0; j < ages.Ncol(i); j++) {
minage = min(ages[i][j], minage);
maxage = max(ages[i][j], maxage);
}
}
found = 0;
for (i = 0; i < stocks.Size(); i++)
if (minage >= stocks[i]->minAge())
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - minimum age less than stock age");
found = 0;
for (i = 0; i < stocks.Size(); i++)
if (maxage <= stocks[i]->maxAge())
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - maximum age greater than stock age");
found = 0;
for (i = 0; i < stocks.Size(); i++)
if (LgrpDiv->maxLength(0) > stocks[i]->getLengthGroupDiv()->minLength())
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - minimum length group less than stock length");
found = 0;
for (i = 0; i < stocks.Size(); i++)
if (LgrpDiv->minLength(LgrpDiv->numLengthGroups()) < stocks[i]->getLengthGroupDiv()->maxLength())
found++;
if (found == 0)
handle.logMessage(LOGWARN, "Warning in catchdistribution - maximum length group greater than stock length");
}
aggregator = new FleetPreyAggregator(fleets, stocks, LgrpDiv, areas, ages, overconsumption);
}
void CatchDistribution::addLikelihood(const TimeClass* const TimeInfo) {
if ((!(AAT.atCurrentTime(TimeInfo))) || (isZero(weight)))
return;
if ((handle.getLogLevel() >= LOGMESSAGE) && ((!yearly) || (TimeInfo->getStep() == TimeInfo->numSteps())))
handle.logMessage(LOGMESSAGE, "Calculating likelihood score for catchdistribution component", this->getName());
int i;
timeindex = -1;
for (i = 0; i < Years.Size(); i++)
if ((Years[i] == TimeInfo->getYear()) && (Steps[i] == TimeInfo->getStep()))
timeindex = i;
if (timeindex == -1)
handle.logMessage(LOGFAIL, "Error in catchdistribution - invalid timestep");
aggregator->Sum();
if ((handle.getLogLevel() >= LOGWARN) && (aggregator->checkCatchData()))
handle.logMessage(LOGWARN, "Warning in catchdistribution - zero catch found");
alptr = &aggregator->getSum();
double l = 0.0;
switch (functionnumber) {
case 1:
l = calcLikMultinomial();
break;
case 2:
l = calcLikPearson(TimeInfo);
break;
case 3:
l = calcLikGamma(TimeInfo);
break;
case 4:
l = calcLikSumSquares(TimeInfo);
break;
case 5:
l = calcLikMVNormal();
break;
case 6:
l = calcLikMVLogistic();
break;
case 7:
l = calcLikLog(TimeInfo);
break;
case 8:
l = calcLikStratified(TimeInfo);
break;
default:
handle.logMessage(LOGWARN, "Warning in catchdistribution - unrecognised function", functionname);
break;
}
if ((!yearly) || (TimeInfo->getStep() == TimeInfo->numSteps())) {
likelihood += l;
if (handle.getLogLevel() >= LOGMESSAGE)
handle.logMessage(LOGMESSAGE, "The likelihood score for this component on this timestep is", l);
}
}
double CatchDistribution::calcLikMultinomial() {
int area, age, len;
int numage = ages.Nrow();
int numlen = LgrpDiv->numLengthGroups();
DoubleVector dist(numage, 0.0);
DoubleVector data(numage, 0.0);
MN.Reset();
//the object MN does most of the work, accumulating likelihood
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = ((*alptr)[area][age][len]).N;
if (numage == 1) {
//only one age-group, so calculate multinomial based on length distribution
likelihoodValues[timeindex][area] +=
MN.calcLogLikelihood((*obsDistribution[timeindex][area])[0],
(*modelDistribution[timeindex][area])[0]);
} else {
//many age-groups, so calculate multinomial based on age distribution per length group
for (len = 0; len < numlen; len++) {
for (age = 0; age < numage; age++) {
dist[age] = (*modelDistribution[timeindex][area])[age][len];
data[age] = (*obsDistribution[timeindex][area])[age][len];
}
likelihoodValues[timeindex][area] += MN.calcLogLikelihood(data, dist);
}
}
}
return MN.getLogLikelihood();
}
double CatchDistribution::calcLikPearson(const TimeClass* const TimeInfo) {
/* written by Hoskuldur Bjornsson 29/8 98
* corrected by kgf 16/9 98
* modified by kgf 11/11 98 to make it possible to sum up catches
* and calculated catches on year basis.
* Modified 3/5 99 by kgf to check the age intervals of the stock
* and the catch data, and make use of the ages that are common
* for the stock and the catch data.*/
double totallikelihood = 0.0;
int age, len, area;
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
//JMB - changed to remove the need to store minrow and mincol stuff ...
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = (*alptr)[area][age][len].N;
if (!yearly) { //calculate likelihood on all steps
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
likelihoodValues[timeindex][area] +=
((*modelDistribution[timeindex][area])[age][len] -
(*obsDistribution[timeindex][area])[age][len]) *
((*modelDistribution[timeindex][area])[age][len] -
(*obsDistribution[timeindex][area])[age][len]) /
fabs(((*modelDistribution[timeindex][area])[age][len] + epsilon));
}
}
totallikelihood += likelihoodValues[timeindex][area];
} else { //calculate likelihood on year basis
if (TimeInfo->getStep() == 1) { //start of a new year
(*modelYearData[area]).setToZero();
(*obsYearData[area]).setToZero();
}
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelYearData[area])[age][len] += (*modelDistribution[timeindex][area])[age][len];
(*obsYearData[area])[age][len] += (*obsDistribution[timeindex][area])[age][len];
}
}
if (TimeInfo->getStep() < TimeInfo->numSteps())
likelihoodValues[timeindex][area] = 0.0;
else { //last step in year, so need to calc likelihood contribution
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
likelihoodValues[timeindex][area] +=
((*modelYearData[area])[age][len] - (*obsYearData[area])[age][len]) *
((*modelYearData[area])[age][len] - (*obsYearData[area])[age][len]) /
fabs(((*modelYearData[area])[age][len] + epsilon));
}
}
totallikelihood += likelihoodValues[timeindex][area];
}
}
}
return totallikelihood;
}
double CatchDistribution::calcLikGamma(const TimeClass* const TimeInfo) {
//written kgf 24/5 00
//Formula by Hans J Skaug, 15/3 00 No weighting at present.
//This function is scale independent.
double totallikelihood = 0.0;
int age, len, area;
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
//JMB - changed to remove the need to store minrow and mincol stuff ...
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = (*alptr)[area][age][len].N;
if (!yearly) { //calculate likelihood on all steps
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
likelihoodValues[timeindex][area] +=
(*obsDistribution[timeindex][area])[age][len] /
((*modelDistribution[timeindex][area])[age][len] + epsilon) +
log((*modelDistribution[timeindex][area])[age][len] + epsilon);
}
}
totallikelihood += likelihoodValues[timeindex][area];
} else { //calculate likelihood on year basis
if (TimeInfo->getStep() == 1) { //start of a new year
(*modelYearData[area]).setToZero();
(*obsYearData[area]).setToZero();
}
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelYearData[area])[age][len] += (*modelDistribution[timeindex][area])[age][len];
(*obsYearData[area])[age][len] += (*obsDistribution[timeindex][area])[age][len];
}
}
if (TimeInfo->getStep() < TimeInfo->numSteps())
likelihoodValues[timeindex][area] = 0.0;
else { //last step in year, so need to calc likelihood contribution
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
likelihoodValues[timeindex][area] +=
(*obsYearData[area])[age][len] / ((*modelYearData[area])[age][len] + epsilon) +
log((*modelYearData[area])[age][len] + epsilon);
}
}
totallikelihood += likelihoodValues[timeindex][area];
}
}
}
return totallikelihood;
}
double CatchDistribution::calcLikLog(const TimeClass* const TimeInfo) {
//written by kgf 23/11 98 to get a better scaling of the stocks.
//modified by kgf 27/11 98 to sum first and then take the logarithm
double totallikelihood = 0.0;
int area, age, len;
double totalmodel, totaldata, ratio;
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
totalmodel = 0.0;
totaldata = 0.0;
//JMB - changed to remove the need to store minrow and mincol stuff ...
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = (*alptr)[area][age][len].N;
if (!yearly) { //calculate likelihood on all steps
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
totalmodel += (*modelDistribution[timeindex][area])[age][len];
totaldata += (*obsDistribution[timeindex][area])[age][len];
}
}
ratio = log(totaldata / totalmodel);
likelihoodValues[timeindex][area] += (ratio * ratio);
} else { //calculate likelihood on year basis
if (TimeInfo->getStep() == 1) { //start of a new year
(*modelYearData[area]).setToZero();
(*obsYearData[area]).setToZero();
}
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelYearData[area])[age][len] += (*modelDistribution[timeindex][area])[age][len];
(*obsYearData[area])[age][len] += (*obsDistribution[timeindex][area])[age][len];
}
}
if (TimeInfo->getStep() < TimeInfo->numSteps())
likelihoodValues[timeindex][area] = 0.0;
else { //last step in year, so need to calculate likelihood contribution
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
totalmodel += (*modelYearData[area])[age][len];
totaldata += (*obsYearData[area])[age][len];
}
}
ratio = log(totaldata / totalmodel);
likelihoodValues[timeindex][area] += (ratio * ratio);
}
}
totallikelihood += likelihoodValues[timeindex][area];
}
return totallikelihood;
}
double CatchDistribution::calcLikSumSquares(const TimeClass* const TimeInfo) {
double temp, totallikelihood, totalmodel, totaldata;
int age, len, area;
totallikelihood = 0.0;
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = ((*alptr)[area][age][len]).N;
totalmodel = 0.0;
totaldata = 0.0;
if (!yearly) { //calculate likelihood on all steps
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
totalmodel += (*modelDistribution[timeindex][area])[age][len];
totaldata += (*obsDistribution[timeindex][area])[age][len];
}
}
if (!(isZero(totalmodel)))
totalmodel = 1.0 / totalmodel;
if (!(isZero(totaldata)))
totaldata = 1.0 / totaldata;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
temp = (((*obsDistribution[timeindex][area])[age][len] * totaldata)
- ((*modelDistribution[timeindex][area])[age][len] * totalmodel));
likelihoodValues[timeindex][area] += (temp * temp);
}
}
totallikelihood += likelihoodValues[timeindex][area];
} else { //calculate likelihood on year basis
if (TimeInfo->getStep() == 1) { //start of a new year
(*modelYearData[area]).setToZero();
(*obsYearData[area]).setToZero();
}
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelYearData[area])[age][len] += (*modelDistribution[timeindex][area])[age][len];
(*obsYearData[area])[age][len] += (*obsDistribution[timeindex][area])[age][len];
}
}
if (TimeInfo->getStep() < TimeInfo->numSteps())
likelihoodValues[timeindex][area] = 0.0;
else { //last step in year, so need to calculate likelihood contribution
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
totalmodel += (*modelYearData[area])[age][len];
totaldata += (*obsYearData[area])[age][len];
}
}
if (!(isZero(totalmodel)))
totalmodel = 1.0 / totalmodel;
if (!(isZero(totaldata)))
totaldata = 1.0 / totaldata;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
temp = (((*obsYearData[area])[age][len] * totaldata)
- ((*modelYearData[area])[age][len] * totalmodel));
likelihoodValues[timeindex][area] += (temp * temp);
}
}
totallikelihood += likelihoodValues[timeindex][area];
}
}
}
return totallikelihood;
}
double CatchDistribution::calcLikStratified(const TimeClass* const TimeInfo) {
int numage = ages.Nrow();
int numlen = LgrpDiv->numLengthGroups();
double temp, totallikelihood, totalmodel, totaldata;
int age, len, area;
totallikelihood = 0.0;
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++)
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
(*modelDistribution[timeindex][area])[age][len] = ((*alptr)[area][age][len]).N;
if (!yearly) { //calculate likelihood on all steps
//calculate an age distribution for each length class
for (len = 0; len < numlen; len++) {
totalmodel = 0.0;
totaldata = 0.0;
for (age = 0; age < numage; age++) {
totalmodel += (*modelDistribution[timeindex][area])[age][len];
totaldata += (*obsDistribution[timeindex][area])[age][len];
}
if (!(isZero(totalmodel)))
totalmodel = 1.0 / totalmodel;
if (!(isZero(totaldata)))
totaldata = 1.0 / totaldata;
for (age = 0; age < numage; age++) {
temp = (((*obsDistribution[timeindex][area])[age][len] * totaldata)
- ((*modelDistribution[timeindex][area])[age][len] * totalmodel));
likelihoodValues[timeindex][area] += (temp * temp);
}
}
totallikelihood += likelihoodValues[timeindex][area];
} else { //calculate likelihood on year basis
if (TimeInfo->getStep() == 1) { //start of a new year
(*modelYearData[area]).setToZero();
(*obsYearData[area]).setToZero();
}
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelYearData[area])[age][len] += (*modelDistribution[timeindex][area])[age][len];
(*obsYearData[area])[age][len] += (*obsDistribution[timeindex][area])[age][len];
}
}
if (TimeInfo->getStep() < TimeInfo->numSteps())
likelihoodValues[timeindex][area] = 0.0;
else { //last step in year, so need to calculate likelihood contribution
//calculate an age distribution for each length class
for (len = 0; len < numlen; len++) {
totalmodel = 0.0;
totaldata = 0.0;
for (age = 0; age < numage; age++) {
totalmodel += (*modelYearData[area])[age][len];
totaldata += (*obsYearData[area])[age][len];
}
if (!(isZero(totalmodel)))
totalmodel = 1.0 / totalmodel;
if (!(isZero(totaldata)))
totaldata = 1.0 / totaldata;
for (age = 0; age < numage; age++) {
temp = (((*obsYearData[area])[age][len] * totaldata)
- ((*modelYearData[area])[age][len] * totalmodel));
likelihoodValues[timeindex][area] += (temp * temp);
}
}
totallikelihood += likelihoodValues[timeindex][area];
}
}
}
return totallikelihood;
}
void CatchDistribution::calcCorrelation() {
int i, j, l, p;
p = LgrpDiv->numLengthGroups();
DoubleMatrix correlation(p, p, 0.0);
for (i = 0; i < lag; i++)
if (fabs(params[i] - 1.0) > 1.0)
illegal = 1;
if (!illegal) {
for (i = 0; i < p; i++) {
for (j = 0; j <= i; j++) {
for (l = 1; l <= lag; l++) {
if ((i - l) >= 0) {
correlation[i][j] += (params[l - 1] - 1.0) * correlation[i - l][j];
correlation[j][i] += (params[l - 1] - 1.0) * correlation[i - l][j];
}
}
}
correlation[i][i] += sigma * sigma;
}
LU = LUDecomposition(correlation);
}
}
double CatchDistribution::calcLikMVNormal() {
double totallikelihood = 0.0;
double sumdata, sumdist;
int age, len, area;
if ((illegal) || (LU.isIllegal()) || isZero(sigma))
return verybig;
DoubleVector diff(LgrpDiv->numLengthGroups(), 0.0);
for (area = 0; area < areas.Nrow(); area++) {
sumdata = 0.0;
sumdist = 0.0;
likelihoodValues[timeindex][area] = 0.0;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelDistribution[timeindex][area])[age][len] = ((*alptr)[area][age][len]).N;
sumdata += (*obsDistribution[timeindex][area])[age][len];
sumdist += (*modelDistribution[timeindex][area])[age][len];
}
}
if (isZero(sumdata))
sumdata = verybig;
else
sumdata = 1.0 / sumdata;
if (isZero(sumdist))
sumdist = verybig;
else
sumdist = 1.0 / sumdist;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
diff.setToZero();
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
diff[len] = ((*obsDistribution[timeindex][area])[age][len] * sumdata)
- ((*modelDistribution[timeindex][area])[age][len] * sumdist);
likelihoodValues[timeindex][area] += diff * LU.Solve(diff);
}
totallikelihood += likelihoodValues[timeindex][area];
}
totallikelihood += LU.getLogDet() * alptr->Size();
return totallikelihood;
}
double CatchDistribution::calcLikMVLogistic() {
double totallikelihood = 0.0;
double sumdata = 0.0, sumdist = 0.0, sumnu = 0.0;
int age, len, area, p;
p = LgrpDiv->numLengthGroups();
DoubleVector nu(p, 0.0);
for (area = 0; area < areas.Nrow(); area++) {
likelihoodValues[timeindex][area] = 0.0;
for (age = (*alptr)[area].minAge(); age <= (*alptr)[area].maxAge(); age++) {
sumdata = 0.0;
sumdist = 0.0;
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
(*modelDistribution[timeindex][area])[age][len] = ((*alptr)[area][age][len]).N;
sumdata += (*obsDistribution[timeindex][area])[age][len];
sumdist += (*modelDistribution[timeindex][area])[age][len];
}
if (isZero(sumdata))
sumdata = verybig;
else
sumdata = 1.0 / sumdata;
if (isZero(sumdist))
sumdist = verybig;
else
sumdist = 1.0 / sumdist;
sumnu = 0.0;
nu.setToZero();
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++) {
nu[len] = log(((*obsDistribution[timeindex][area])[age][len] * sumdata) + verysmall)
- log(((*modelDistribution[timeindex][area])[age][len] * sumdist) + verysmall);
sumnu += nu[len];
}
sumnu = sumnu / p;
for (len = (*alptr)[area].minLength(age); len < (*alptr)[area].maxLength(age); len++)
likelihoodValues[timeindex][area] += (nu[len] - sumnu) * (nu[len] - sumnu);
}
totallikelihood += likelihoodValues[timeindex][area];
}
if (isZero(sigma)) {
handle.logMessage(LOGWARN, "Warning in catchdistribution - multivariate logistic sigma is zero");
return verybig;
}
totallikelihood = (totallikelihood / (sigma * sigma)) + (log(sigma) * (p - 1));
return totallikelihood;
}
void CatchDistribution::printSummary(ofstream& outfile) {
int year, area;
for (year = 0; year < likelihoodValues.Nrow(); year++) {
for (area = 0; area < likelihoodValues.Ncol(year); area++) {
if (!yearly) {
outfile << setw(lowwidth) << Years[year] << sep << setw(lowwidth)
<< Steps[year] << sep << setw(printwidth) << areaindex[area] << sep
<< setw(largewidth) << this->getName() << sep << setw(smallwidth) << weight
<< sep << setprecision(largeprecision) << setw(largewidth)
<< likelihoodValues[year][area] << endl;
} else {
if (isZero(likelihoodValues[year][area])) {
// assume that this isnt the last step for that year and ignore
} else {
outfile << setw(lowwidth) << Years[year] << " all "
<< setw(printwidth) << areaindex[area] << sep
<< setw(largewidth) << this->getName() << sep << setprecision(smallprecision)
<< setw(smallwidth) << weight << sep << setprecision(largeprecision)
<< setw(largewidth) << likelihoodValues[year][area] << endl;
}
}
}
}
outfile.flush();
}