Skip to content

Updated I5 processes (reg_income). #264

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
dav-sonn merged 1 commit into from
Oct 8, 2025
Merged

Updated I5 processes (reg_income). #264

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 14 additions & 12 deletions src/main/java/simpaths/data/Parameters.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -527,11 +527,13 @@ else if(numberOfChildren <= 5) {
private static MultiKeyCoefficientMap coeffCovarianceIncomeI3a; //Capital income if in continuous education
private static MultiKeyCoefficientMap coeffCovarianceIncomeI3b; //Capital income if not in continuous education
private static MultiKeyCoefficientMap coeffCovarianceIncomeI3c; //Pension income for those aged over 50 who are not in continuous education
private static MultiKeyCoefficientMap coeffCovarianceIncomeI4a, coeffCovarianceIncomeI4b; // Pension income for those moving from employment to retirement (I4a) and those already retired (I4b)
private static MultiKeyCoefficientMap coeffCovarianceIncomeI5a_selection, coeffCovarianceIncomeI5b_amount; // Selection equation for receiving pension income for those moving from employment to retirement (I5a) and amount in levels (I5b)
private static MultiKeyCoefficientMap coeffCovarianceIncomeI4a;
private static MultiKeyCoefficientMap coeffCovarianceIncomeI4b; //Pension income for those already retired
private static MultiKeyCoefficientMap coeffCovarianceIncomeI5a; //Pension income for those moving from employment to retirement
private static MultiKeyCoefficientMap coeffCovarianceIncomeI6a_selection, coeffCovarianceIncomeI6b_amount; // Selection equation for receiving pension income for those in retirement (I6a) and amount in levels (I6b), in the initial simulated year
private static MultiKeyCoefficientMap coeffCovarianceIncomeI3a_selection; //Probability of receiving capital income if in continuous education
private static MultiKeyCoefficientMap coeffCovarianceIncomeI3b_selection; //Probability of receiving capital income if not in continuous education
private static MultiKeyCoefficientMap coeffCovarianceIncomeI5a_selection; //Selection equation for receiving pension income for those moving from employment to retirement

//Homeownership
private static MultiKeyCoefficientMap coeffCovarianceHomeownership; //Probit regression assigning homeownership status
Expand Down Expand Up @@ -765,7 +767,7 @@ else if(numberOfChildren <= 5) {
private static LinearRegression regIncomeI3c;
private static LinearRegression regIncomeI4a;
private static LinearRegression regIncomeI4b;
private static LinearRegression regIncomeI5b_amount;
private static LinearRegression regIncomeI5a;
private static LinearRegression regIncomeI6b_amount;
private static BinomialRegression regIncomeI3a_selection;
private static BinomialRegression regIncomeI3b_selection;
Expand Down Expand Up @@ -1047,11 +1049,11 @@ public static void loadParameters(Country country, int maxAgeModel, boolean enab
int columnsIncomeI4a = -1;
int columnsIncomeI4b = -1;
int columnsIncomeI5a = -1;
int columnsIncomeI5b = -1;
int columnsIncomeI6a = -1;
int columnsIncomeI6b = -1;
int columnsIncomeI3a_selection = -1;
int columnsIncomeI3b_selection = -1;
int columnsIncomeI5a_selection = -1;
int columnsLeaveHomeP1a = -1;
int columnsHomeownership = -1;
int columnsRetirementR1a = -1;
Expand Down Expand Up @@ -1205,12 +1207,12 @@ else if(country.equals(Country.UK)) {
columnsIncomeI3c = 28; //*
columnsIncomeI4a = 24; //*
columnsIncomeI4b = 31;
columnsIncomeI5a = 30;
columnsIncomeI5b = 31;
columnsIncomeI5a = 31;
columnsIncomeI6a = 22; //*
columnsIncomeI6b = 22; //*
columnsIncomeI3a_selection = 26;
columnsIncomeI3b_selection = 36;
columnsIncomeI5a_selection = 30;
columnsLeaveHomeP1a = 25;
columnsHomeownership = 41;
columnsRetirementR1a = 33;
Expand Down Expand Up @@ -1428,10 +1430,10 @@ else if (country.equals(Country.IT)) {
coeffCovarianceIncomeI3a = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I3a_amount", 1, columnsIncomeI3a);
coeffCovarianceIncomeI3b = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I3b_amount", 1, columnsIncomeI3b);
coeffCovarianceIncomeI4b = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I4b_amount", 1, columnsIncomeI4b);
coeffCovarianceIncomeI5a_selection = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I5a_selection", 1, columnsIncomeI5a);
coeffCovarianceIncomeI5b_amount = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I5a_amount", 1, columnsIncomeI5b);
coeffCovarianceIncomeI5a = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I5a_amount", 1, columnsIncomeI5a);
coeffCovarianceIncomeI3a_selection = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I3a_selection", 1, columnsIncomeI3a_selection);
coeffCovarianceIncomeI3b_selection = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I3b_selection", 1, columnsIncomeI3b_selection);
coeffCovarianceIncomeI5a_selection = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_income.xlsx", countryString + "_I5a_selection", 1, columnsIncomeI5a_selection);

//Leaving parental home
coeffCovarianceLeaveHomeP1a = ExcelAssistant.loadCoefficientMap(Parameters.getInputDirectory() + "reg_leaveParentalHome.xlsx", countryString + "_P1a", 1, columnsLeaveHomeP1a);
Expand Down Expand Up @@ -1525,12 +1527,12 @@ else if (country.equals(Country.IT)) {
//coeffCovarianceIncomeI3c = RegressionUtils.bootstrap(coeffCovarianceIncomeI3c);
//coeffCovarianceIncomeI4a = RegressionUtils.bootstrap(coeffCovarianceIncomeI4a);
coeffCovarianceIncomeI4b = RegressionUtils.bootstrap(coeffCovarianceIncomeI4b);
coeffCovarianceIncomeI5a_selection = RegressionUtils.bootstrap(coeffCovarianceIncomeI5a_selection);
coeffCovarianceIncomeI5b_amount = RegressionUtils.bootstrap(coeffCovarianceIncomeI5b_amount);
coeffCovarianceIncomeI5a = RegressionUtils.bootstrap(coeffCovarianceIncomeI5a);
//coeffCovarianceIncomeI6a_selection = RegressionUtils.bootstrap(coeffCovarianceIncomeI6a_selection);
//coeffCovarianceIncomeI6b_amount = RegressionUtils.bootstrap(coeffCovarianceIncomeI6b_amount);
coeffCovarianceIncomeI3a_selection = RegressionUtils.bootstrap(coeffCovarianceIncomeI3a_selection);
coeffCovarianceIncomeI3b_selection = RegressionUtils.bootstrap(coeffCovarianceIncomeI3b_selection);
coeffCovarianceIncomeI5a_selection = RegressionUtils.bootstrap(coeffCovarianceIncomeI5a_selection);

//Leave parental home
coeffCovarianceLeaveHomeP1a = RegressionUtils.bootstrap(coeffCovarianceLeaveHomeP1a);
Expand Down Expand Up @@ -1649,7 +1651,7 @@ else if (country.equals(Country.IT)) {
//regIncomeI3c = new LinearRegression(coeffCovarianceIncomeI3c);
//regIncomeI4a = new LinearRegression(coeffCovarianceIncomeI4a);
regIncomeI4b = new LinearRegression(coeffCovarianceIncomeI4b);
regIncomeI5b_amount = new LinearRegression(coeffCovarianceIncomeI5b_amount);
regIncomeI5a = new LinearRegression(coeffCovarianceIncomeI5a);
//regIncomeI6b_amount = new LinearRegression(coeffCovarianceIncomeI6b_amount);
regIncomeI3a_selection = new BinomialRegression(RegressionType.Logit, Indicator.class, coeffCovarianceIncomeI3a_selection);
regIncomeI3b_selection = new BinomialRegression(RegressionType.Logit, Indicator.class, coeffCovarianceIncomeI3b_selection);
Expand Down Expand Up @@ -2151,7 +2153,7 @@ public static void setEmploymentsFurloughedFlex(MultiKeyCoefficientMap employmen
public static LinearRegression getRegIncomeI3c() { return regIncomeI3c; }
public static LinearRegression getRegIncomeI4a() { return regIncomeI4a; }
public static LinearRegression getRegIncomeI4b() {return regIncomeI4b;}
public static LinearRegression getRegIncomeI5b_amount() {return regIncomeI5b_amount;}
public static LinearRegression getRegIncomeI5a() {return regIncomeI5a;}
public static LinearRegression getRegIncomeI6b_amount() { return regIncomeI6b_amount; }
public static BinomialRegression getRegIncomeI3a_selection() { return regIncomeI3a_selection; }
public static BinomialRegression getRegIncomeI3b_selection() { return regIncomeI3b_selection; }
Expand Down
8 changes: 4 additions & 4 deletions src/main/java/simpaths/model/Person.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1720,7 +1720,7 @@ protected void updateNonLabourIncome() {

Private pension income when individual moves from non-retirement to retirement is modelled using:
i) process I5a_selection, to determine who receives private pension income
ii) process I5b_amount, for those who are determined to receive private pension income by process I5a_selection. I5b_amount is modelled in levels using linear regression.
ii) process I5a_amount, for those who are determined to receive private pension income by process I5a_selection. I5a_amount is modelled in levels using linear regression.
*/

double score, rmse, pensionIncLevel = 0.;
Expand All @@ -1731,14 +1731,14 @@ protected void updateNonLabourIncome() {
rmse = Parameters.getRMSEForRegression("I4b");
pensionIncLevel = setIncomeBySource(score, rmse, IncomeSource.PrivatePension, RegressionScoreType.Asinh);
} else {
// For individuals in the first year of retirement, use processes I5a_selection and I5b_amount
// For individuals in the first year of retirement, use processes I5a_selection and I5a

double prob = Parameters.getRegIncomeI5a_selection().getProbability(this, Person.DoublesVariables.class);
boolean hasPrivatePensionIncome = (innovations.getDoubleDraw(19) < prob);
if (hasPrivatePensionIncome) {

score = Parameters.getRegIncomeI5b_amount().getScore(this, Person.DoublesVariables.class);
rmse = Parameters.getRMSEForRegression("I5b");
score = Parameters.getRegIncomeI5a().getScore(this, Person.DoublesVariables.class);
rmse = Parameters.getRMSEForRegression("I5a");
pensionIncLevel = setIncomeBySource(score, rmse, IncomeSource.PrivatePension, RegressionScoreType.Level);
}
}
Expand Down