It not working. how do i debug to find out error

Question

It not working. how do i debug to find out error

parveenmani opened this issue 9 years ago · 11 comments

optimizer.optif9_f77(n, x, minclass, typsiz, fscale, method, iexp, msg, ndigit, itnlim,
iagflg, iahflg, dlt, gradtl, stepmx, steptl, xpls, fpls, gpls, itrmcd, a, udiag);

is notworking

Answer 1 · 2015-07-09T12:22:54.000Z

This code is a Java translation of a FORTRAN program. I didn't write the program. If you include an example of how it does not work, I will help you trouble shoot the problem. However, consider using DefaultUncminOptimizer.java instead. It provides a better interface to optimizer.optif9_f77.

Answer 2 · 2015-07-11T08:31:16.000Z

Sir,
i am thankful to you for your reply
can you please provide the format of input file given to IRT project.
i want to give value manually. in short i want to know where should i enter
value.
presently i have made some changes in "ItemResponseFileSummary.java"
i am giving value to line[0]=1 line[2]=1
i just wanted find out value of theta by providing level of difficulty and
is correct

On Thu, Jul 9, 2015 at 5:52 PM, Patrick Meyer notifications@github.com
wrote:

This code is a Java translation of a FORTRAN program. I didn't write the
program. If you include an example of how it does not work, I will help you
trouble shoot the problem. However, consider using
DefaultUncminOptimizer.java instead. It provides a better interface to
optimizer.optif9_f77.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 3 · 2015-07-11T18:10:40.000Z

Please review the test cases to learn how to use the classes. For item calibration, see the MarginalMaximumLiklihoodEstimationTest.java class and for person scoring see IrtExamineeTest.java. They both can be found in https://github.com/meyerjp3/psychometrics/tree/master/src/test/java/com/itemanalysis/psychometrics/irt/estimation

Answer 4 · 2015-07-12T10:51:45.000Z

Sir,
What does this item stands for?? is it difficulty?
I want to give following inputs
Question is correct or not (1/0)
Difficulty (1 to 5)
Ability of a person
Depending upon this value i wanted to calculate or update ability of a
person. in such scenario which classes should i mainly use.

one more query. if their are 5 types of difficulty leven then their would
be 5 types of ability of a person. M i correct??

On Sat, Jul 11, 2015 at 11:40 PM, Patrick Meyer notifications@github.com
wrote:

Please review the test cases to learn how to use the classes. For item
calibration, see the MarginalMaximumLiklihoodEstimationTest.java class and
for person scoring see IrtExamineeTest.java. They both can be found in
https://github.com/meyerjp3/psychometrics/tree/master/src/test/java/com/itemanalysis/psychometrics/irt/estimation

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 5 · 2015-07-12T14:49:13.000Z

The items are the item responses. O or 1 for binary items. 0, 1, 2, and so on for polytomous items.

Answer 6 · 2015-07-12T15:31:21.000Z

sir,
can you please help me out.
i need to create an interface which basically takes two inputs

is correct (answer is correct or not means 0 or 1)
difficulty level (we have 5 difficulty level from 1 to 5)
and i have to find out ability/theta using these inputs. i am already
reading all classes but time is a constraint.
can you please guide which classes should i concentrate on to achieve my
target.

i would be thankful and it would be a great help.

Regards
Ms Parveen Sharma

On Sun, Jul 12, 2015 at 8:19 PM, Patrick Meyer notifications@github.com
wrote:

The items are the item responses. O or 1 for binary items. 0, 1, 2, and so
on for polytomous items.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 7 · 2015-07-12T16:08:40.000Z

How did you obtain the difficulty estimates and why are they ordinal? Why not estimate the item parameters and theta simultaneously instead of constraining the difficulty parameters to values a priori? Maybe if you are able to provide a bit more context (e.g., purpose, types of items/response scales, how the discrimination parameter is estimated and/or why there are difficulty parameter estimates without discrimination parameter estimates, etc...) Dr Meyer and/or other users following the repository might be able to provide you with answers that are more relevant to your particular context.

Answer 8 · 2015-07-13T12:45:20.000Z

Sir,
As per your Guidance i have calculated theta from "IrtExaminee (and method
pcfEstimate ) " am i correct or doing wrong
but i have a doubt it would be correct or not.
what i am doing is :
passing binary value to Line[] MMLESimulator,then i am calling

String[] line = new String[5];
ArrayList responseData = new
ArrayList();
byte[] rv = null;
ItemResponseVector irv = null;
ItemResponseVector[] itemResponseVector = null;
line[0]="1";
line[1]="0";
line[2]="0";
line[3]="0";
line[4]="0";

            rv=new byte[5];

            for(int j=0;j<5;j++)

            {

                rv[j] = Byte.parseByte(line[j]);

            }
            irv = new ItemResponseVector(rv, 1.0);
            responseData.add(irv);


        itemResponseVector = new

ItemResponseVector[responseData.size()];

        for(int i=0;i<responseData.size();i++){
            itemResponseVector[i] = responseData.get(i);

        }


        try{
        ItemResponseModel irm=new ItemResponseModel();
    IrtExaminee ie= new IrtExaminee(irm,itemResponseVector);
        }
        catch(DimensionMismatchException ex)
        {
        ex.printStackTrace();

        }

and then using pcfEstimate to get theta value....
please guide???

On Sun, Jul 12, 2015 at 9:38 PM, William Buchanan notifications@github.com
wrote:

How did you obtain the difficulty estimates and why are they ordinal? Why
not estimate the item parameters and theta simultaneously instead of
constraining the difficulty parameters to values a priori? Maybe if you are
able to provide a bit more context (e.g., purpose, types of items/response
scales, how the discrimination parameter is estimated and/or why there are
difficulty parameter estimates without discrimination parameter estimates,
etc...) Dr Meyer and/or other users following the repository might be able
to provide you with answers that are more relevant to your particular
context.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 9 · 2015-07-15T08:53:42.000Z

Sir,
Which class has connection between ability and difficulty.
if one answer first question correct that what would be value of theta.?? i
just want to cross check the value

On Mon, Jul 13, 2015 at 6:15 PM, Parveen Mani parveen@edulabs.co.in wrote:

Sir,
As per your Guidance i have calculated theta from "IrtExaminee (and
method pcfEstimate ) " am i correct or doing wrong
but i have a doubt it would be correct or not.
what i am doing is :
passing binary value to Line[] MMLESimulator,then i am calling

String[] line = new String[5];
ArrayList responseData = new
ArrayList();
byte[] rv = null;
ItemResponseVector irv = null;
ItemResponseVector[] itemResponseVector = null;
line[0]="1";
line[1]="0";
line[2]="0";
line[3]="0";
line[4]="0";
            rv=new byte[5];

            for(int j=0;j<5;j++)

            {

                rv[j] = Byte.parseByte(line[j]);

            }
            irv = new ItemResponseVector(rv, 1.0);
            responseData.add(irv);


        itemResponseVector = new
ItemResponseVector[responseData.size()];
        for(int i=0;i<responseData.size();i++){
            itemResponseVector[i] = responseData.get(i);

        }


        try{
        ItemResponseModel irm=new ItemResponseModel();
    IrtExaminee ie= new IrtExaminee(irm,itemResponseVector);
        }
        catch(DimensionMismatchException ex)
        {
        ex.printStackTrace();

        }
and then using pcfEstimate to get theta value....
please guide???

On Sun, Jul 12, 2015 at 9:38 PM, William Buchanan <
notifications@github.com> wrote:

How did you obtain the difficulty estimates and why are they ordinal? Why
not estimate the item parameters and theta simultaneously instead of
constraining the difficulty parameters to values a priori? Maybe if you are
able to provide a bit more context (e.g., purpose, types of items/response
scales, how the discrimination parameter is estimated and/or why there are
difficulty parameter estimates without discrimination parameter estimates,
etc...) Dr Meyer and/or other users following the repository might be able
to provide you with answers that are more relevant to your particular
context.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 10 · 2015-07-16T07:02:41.000Z

Sir,
I am using IrtExaminee to calculate theta , how can i use adjustment and
xmin and xmax.
i have response as binary i.e 0 for incorrect and 1 for incorrect

On Wed, Jul 15, 2015 at 2:23 PM, Parveen Mani parveen@edulabs.co.in wrote:

Sir,
Which class has connection between ability and difficulty.
if one answer first question correct that what would be value of theta.??
i just want to cross check the value

On Mon, Jul 13, 2015 at 6:15 PM, Parveen Mani parveen@edulabs.co.in
wrote:
Sir,
As per your Guidance i have calculated theta from "IrtExaminee (and
method pcfEstimate ) " am i correct or doing wrong
but i have a doubt it would be correct or not.
what i am doing is :
passing binary value to Line[] MMLESimulator,then i am calling

String[] line = new String[5];
ArrayList responseData = new
ArrayList();
byte[] rv = null;
ItemResponseVector irv = null;
ItemResponseVector[] itemResponseVector = null;
line[0]="1";
line[1]="0";
line[2]="0";
line[3]="0";
line[4]="0";
            rv=new byte[5];

            for(int j=0;j<5;j++)

            {

                rv[j] = Byte.parseByte(line[j]);

            }
            irv = new ItemResponseVector(rv, 1.0);
            responseData.add(irv);


        itemResponseVector = new
ItemResponseVector[responseData.size()];
        for(int i=0;i<responseData.size();i++){
            itemResponseVector[i] = responseData.get(i);

        }


        try{
        ItemResponseModel irm=new ItemResponseModel();
    IrtExaminee ie= new IrtExaminee(irm,itemResponseVector);
        }
        catch(DimensionMismatchException ex)
        {
        ex.printStackTrace();

        }
and then using pcfEstimate to get theta value....
please guide???

On Sun, Jul 12, 2015 at 9:38 PM, William Buchanan <
notifications@github.com> wrote:

How did you obtain the difficulty estimates and why are they ordinal?
Why not estimate the item parameters and theta simultaneously instead of
constraining the difficulty parameters to values a priori? Maybe if you are
able to provide a bit more context (e.g., purpose, types of items/response
scales, how the discrimination parameter is estimated and/or why there are
difficulty parameter estimates without discrimination parameter estimates,
etc...) Dr Meyer and/or other users following the repository might be able
to provide you with answers that are more relevant to your particular
context.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

Answer 11 · 2015-07-16T13:07:26.000Z

The best thing to do is study the test cases. Also, considering reading more about person scoring in item response theory. Details of the joint maximum likelihood estimation algorithm may be found in Meyer & Hailey (2012) , http://europepmc.org/abstract/MED/23234828 The book Applied Measurement with jMetrik also provide more details about the procedures in the psychometrics library. Finally, read the API, http://www.itemanalysis.com/api/psychometrics/. It is a little out of date, but it contains the information you need to answer your questions.