Skip to content

Issue with the base model at the predition step with AALASSO #281

@MorganePhilipp

Description

@MorganePhilipp

Hello,

I tried to use the adaptive adjusted lasso (AALASSO) in PsN, but I encountered some issues.
For t = 0, i.e., with the base model without any covariates, it is impossible to obtain results for the prediction models. I got the following error message:

Parsing relations and calculating covariate statistics
 ... done
Running the basic xv_step (option pred_ofv_start_t undefined).
Starting 5 NONMEM executions. 5 in parallel.
S:1 .. S:2 .. S:3 .. S:4 .. S:5 .. Job <35692572> is not found
lsf run error, jobID 35692572 not recognized by system:
Waiting for all NONMEM runs to finish:
F:1 .. Job <35692574> is not found
lsf run error, jobID 35692574 not recognized by system:
F:2 .. Job <35692575> is not found
lsf run error, jobID 35692575 not recognized by system:
F:3 .. Job <35692576> is not found
lsf run error, jobID 35692576 not recognized by system:
F:4 .. Job <35692577> is not found
lsf run error, jobID 35692577 not recognized by system:
F:5 ..  done

I tried to launch my model with clean = 1 to better understand what was going on, but I couldn’t spot any problem in the created NONMEM files.

In the raw_results.csv file located in lasso_dir4/basic_xv_step/prediction_dir1/, I get the following output:

"model","problem","subproblem","covariance_step_run","minimization_successful","covariance_step_successful","covariance_step_warnings","estimate_near_boundary","rounding_errors","zero_gradients","final_zero_gradients","hessian_reset","s_matrix_singular","significant_digits","condition_number","est_methods","model_run_time","subprob_est_time","subprob_cov_time","ofv","KA","CL","V2","Q","V3","PROP_ERR","ADD_ERR","IIV_KA","IIV_CL","IIV_V2","IIV_Q","IIV_V3","RES_VAR","seKA","seCL","seV2","seQ","seV3","sePROP_ERR","seADD_ERR","seIIV_KA","seIIV_CL","seIIV_V2","seIIV_Q","seIIV_V3","seRES_VAR","shrinkage_eta1(%)","shrinkage_eta2(%)","shrinkage_eta3(%)","shrinkage_eta4(%)","shrinkage_eta5(%)","shrinkage_iwres(%)","EI1","EI2","EI3","EI4","EI5","EI6","EI7","EI8","EI9","EI10","EI11","EI12","EI13"
1,1,1,run failed: NONMEM run failed
2,1,1,run failed: NONMEM run failed
3,1,1,run failed: NONMEM run failed
4,1,1,run failed: NONMEM run failed
5,1,1,run failed: NONMEM run failed

On the other hand, the estimation part for t = 0 was successful.

Also, using the same model file, I tried running ALASSO and LASSO without any issues.

Any idea what might be going wrong?
Just to let you know, I’m using PsN version 5.4.0 and NONMEM version 7.5.1.

Thank you so much for your help,
Morgane

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions