topic Re: JMP script to generate cumulative value after modeling in Discussions

JMP script to generate cumulative value after modeling

Caozheng0115 — Thu, 28 Aug 2025 14:11:17 GMT

Hi, all
I am new to JMP and tried to write a script
I use JMP 18 standard and perform recurrence analysis. The model I use is loglinear NHPP. I want to generate specific cumulative value at 100k, 200k, 300k, 400k and 500k transactions. I can see there is click option in the UI for each transaction (shown in the attachment). But I want to automate this using scripts. Following is the script I tried: It generate an error in the figure. Is there Could you help me with the code?

// Run the analysis and capture the report
rec_report = Recurrence Analysis(
	Y( :Accumulated Transactions ),
	Label( :BOT ID ),
	Cost( :SAP Purchase Price ),
	Grouping( :Bot Series ),
	Event Plot( 0 ),
	MCF Confid Limits( 1 ),
	Fit Model(
		Scale Effects( :Bot Series ),
		Shape Effects( :Bot Series ),
		Run Model,
		Model Type( "Loglinear Nonhomogeneous Poisson Process" )
	)
);

// --- Extract Estimates ---
// Get the "Estimates" table directly from the report and convert it to a scriptable object
estimates_table = rec_report["Loglinear NHPP Report", "Estimates"] << Table Box;

// Get the 'Estimate' column values as a list.
// The order is consistently Lambda, then Beta.
estimates_list = estimates_table << Get Values( "Estimate" );

// Assign the parameter estimates from the list
lambda_hat = estimates_list[1]; // Lambda (λ) is the first estimate
beta_hat = estimates_list[2]; // Beta (β) is the second estimate

// Define the transactions for which to calculate values
transactions = {100000, 200000, 300000, 400000, 500000};

// Calculate and display intensity and cumulative values
For( i = 1, i <= N Items( transactions ), i++,
	transaction_time = transactions[i];

	// Cumulative Mean Function C(t) = λ * t^β
	cumulative = lambda_hat * (transaction_time ^ beta_hat);

	// Instantaneous Intensity λ(t) = λ * β * t^(β-1)
	intensity = lambda_hat * beta_hat * (transaction_time ^ (beta_hat - 1));

	// Print the results to the log
	Show( transaction_time, cumulative, intensity );
);

Re: JMP script to generate cumulative value after modeling

jthi — Thu, 28 Aug 2025 14:30:08 GMT

Most likely you wish to include Table Box inside your report subscript instead of sending a message to it

estimates_table = rec_report["Loglinear NHPP Report", "Estimates", Table Box(1)]

Also I think your "Estimates" should be "Parameter Estimates" (or use ? wildcard)'. Here is example until that point using JMP's sample data

Names Default To Here(1);
dt = Open("$SAMPLE_DATA/Reliability/Bladder Cancer.jmp");

ra = dt << Recurrence Analysis(
	Y(:Age),
	Label(:Patient Number),
	Cost(:Cost),
	Grouping(:Treatment Group),
	Event Plot(1),
	Fit Model(
		Scale Effects(:Intensity),
		Shape Effects(:Intensity),
		Run Model,
		Model Type("Power Nonhomogeneous Poisson Process")
	)
);


tb = Report(ra)[OutlineBox("Fitted Recurrence Model"), OutlineBox("Parameter Estimates"), TableBox(1)];
estimates = tb[NumberColBox("Estimate")] << get;

Re: JMP script to generate cumulative value after modeling

Caozheng0115 — Fri, 29 Aug 2025 15:01:55 GMT

Hi, jthi

It works now. Thanks a lot.
I should not try to create my own formular. There is 'Specific Intensity and Cumulative' under 'Fitted Recurrence Model' after I run the model in JMP. It can generate the cumulative value with confidence interval. I can just use the formular instead of creating a new one. I tried to get the formular to generate new results in a table. But failed. Could you help? The error is in the image. Thanks again.

Names Default To Here( 1 );

// --- Run the analysis ---
rec_report = Recurrence Analysis(
	Y( :Accumulated Transactions ),
	Label( :BOT ID ),
	Cost( :SAP Purchase Price ),
	Grouping( :Bot Series ),
	Event Plot( 0 ),
	MCF Confid Limits( 1 ),
	Fit Model(
		Scale Effects( :Bot Series ),
		Shape Effects( :Bot Series ),
		Run Model,
		Model Type( "Loglinear Nonhomogeneous Poisson Process" )
	)
);

// --- Save JMP’s own formulas into the source table ---
rec_report << Save Intensity Formula;
rec_report << Save Cumulative Formula;

// Grab those formulas so we can reuse them in a clean table
dt_src=Current Data Table();
intensity_formula  = Column( dt_src, "Intensity" )  << Get Formula;
cumulative_formula = Column( dt_src, "Cumulative" ) << Get Formula;

// --- Define the transaction times you care about ---
transactions = {100000, 200000, 300000, 400000, 500000};

// --- Build a results table with JMP-calculated fitted values ---
dt_sic = New Table( "Specific Intensity and Cumulative",
	New Column( "Accumulated Transactions", Numeric, "Continuous", Values( transactions ) ),
	New Column( "Intensity",  Numeric, Formula( intensity_formula ) ),
	New Column( "Cumulative", Numeric, Formula( cumulative_formula ) )
);

// Evaluate formulas and show the table
dt_sic << Recalculate;
dt_sic << Bring Window To Front;

Re: JMP script to generate cumulative value after modeling

jthi — Fri, 29 Aug 2025 17:15:52 GMT

Your script very strongly looks like AI generated... Note that these are not documented in Scripting Index (<< Save Intensity Formula and << Save Cumulative Formula) but you can "figure it out" by tinkering around

Names Default To Here(1);
dt = Open("$SAMPLE_DATA/Reliability/Bladder Cancer.jmp");
ra = dt << Recurrence Analysis(
	Y(:Age),
	Label(:Patient Number),
	Cost(:Cost),
	Grouping(:Treatment Group),
	Event Plot(1),
	Fit Model(
		Scale Effects(:Age),
		Run Model,
		Model Type("Proportional Intensity Poisson Process")
	)
);

fit = Report(ra)[OutlineBox("Fitted Recurrence Model")] << Get Scriptable Object;

intensity_col = fit << Save Intensity Formula;
cumul_col = fit << Save Cumulative Formula;

i_f = intensity_col << Get Formula;
c_f = cumul_col << Get Formula;

transactions = {1, 2, 3, 4, 5};
dt_s = New Table("Specific Intensity and Cumulative",
	New Column("Age", Numeric, "Continuous", Values(transactions))
);

Eval(Eval Expr(
	dt_s << New Column("Intensity", Numeric, Continuous, Formula(
		Expr(NameExpr(i_f))
	));
));

Eval(Eval Expr(
	dt_s << New Column("Cumulative", Numeric, Continuous, Formula(
		Expr(NameExpr(c_f))
	));
));

Write();

@jules

Re: JMP script to generate cumulative value after modeling

Caozheng0115 — Fri, 29 Aug 2025 18:29:05 GMT

Hi, jthi
Thanks, this works. All my code is generated by AI. I start to script in JMP last week. I can hardly wrote a single line of code by myself. Is there a place that I can learn JMP script?
Thanks.

Re: JMP script to generate cumulative value after modeling

jthi — Fri, 29 Aug 2025 19:12:02 GMT

Use enhanced log in JMP and learn the different methods how JMP is able to create some of the scripts for you. There is also Scripting Guide , you have direct access to Scripting Index from JMP's Help menu and JMP Community does have a lot of content like Introduction to the JMP Scripting Language , JMP Scripters Club and JSL Cookbook .