29913106cb
Add get_sequence_lengths method to HealthDataset and create extract_sequence_lengths script for per-patient sequence length extraction
2026-01-19 14:39:13 +08:00
76d3fed76f
Enhance compute_disease_capture_at_k_fast to support return counts and update LandmarkEvaluator for backward compatibility with new metrics structure
2026-01-19 00:01:21 +08:00
d13fa430b7
Validate bin edges for Piecewise Exponential CIF Loss to ensure at least two finite edges and correct starting point
2026-01-18 20:54:47 +08:00
6de2132e84
Add fast vectorized implementation for Disease-Capture@K and enhance LandmarkEvaluator with profiling and correctness check options
2026-01-18 18:14:45 +08:00
6e76d67a10
Enhance _maybe_torch_compile and add _maybe_cudagraph_mark_step_begin for improved CUDA Graphs handling
2026-01-18 18:04:54 +08:00
a4b19b6e08
Enhance LandmarkEvaluator with model compilation and optimization options
2026-01-18 17:56:59 +08:00
014393a33f
Add support for evaluation indices in LandmarkEvaluator class
2026-01-18 17:43:47 +08:00
0057bc0dd9
Refactor evaluation scripts for multi-GPU execution
...
- Removed `run_evaluations_multi_gpu.sh` script as it was redundant.
- Updated `run_experiments_multi_gpu.sh` to handle evaluation jobs instead of training.
- Changed command-line options to support evaluation-specific parameters.
- Implemented run directory discovery and validation for evaluation jobs.
- Enhanced logging to capture evaluation details and outputs.
- Added options for centralized output management and skipping existing results.
2026-01-18 17:38:20 +08:00
b80d9a4256
Remove evaluate_next_event.py and utils.py files to streamline the codebase. These files contained functions and classes related to evaluation and utility operations that are no longer needed.
2026-01-18 15:41:07 +08:00
1d682c130d
Implement control for implicit parallelism using --max_cpu_cores argument
2026-01-18 00:06:33 +08:00
cfe7f88162
Add --max_cpu_cores argument for parallel processing in evaluation scripts
2026-01-17 23:53:24 +08:00
248fb09c34
Refactor run_evaluations_multi_gpu.sh to remove --tau-short requirement and add support for next-args and horizon-args files
2026-01-17 23:39:55 +08:00
4686b56336
Refactor next-event evaluation to use next-token scores and implement clean control AUC metrics
2026-01-17 23:34:19 +08:00
e56068e668
Add compare_models.py for model evaluation with primary metrics and age-bin validation
2026-01-17 16:12:56 +08:00
fcd948818c
Enhance next-event evaluation with age-bin metrics and diagnostic AUC outputs
2026-01-17 15:31:12 +08:00
197842b1a6
Add script for multi-GPU evaluations with flexible options and logging
2026-01-17 14:56:45 +08:00
a90f22a865
Enhance HealthDataset with caching for event tensors and update evaluation scripts to use test subsets
2026-01-17 14:42:02 +08:00
7840a4c53e
Refactor next-event evaluation logic and add age-bin metrics output
2026-01-17 14:09:50 +08:00
67f92ce6c4
Add tqdm progress bar support and disable option for evaluation scripts
2026-01-17 14:00:42 +08:00
bfab601a77
Add evaluation scripts for next-event prediction and horizon-capture evaluation with detailed metric disclaimers
2026-01-17 13:49:39 +08:00
07916ee529
Remove torch_metrics.py file to streamline evaluation metrics implementation
2026-01-17 13:14:43 +08:00
c1bba30de4
Remove evaluation_age_time_dependent.py and utils.py files
...
- Deleted the entire evaluation_age_time_dependent.py file which contained functions for evaluating age-dependent metrics, including various statistical calculations and data aggregation methods.
- Removed utils.py file that provided utility functions for sampling context in fixed age bins and multi-hot encoding for disease occurrences.
2026-01-17 09:53:15 +08:00
a637beb220
Add function to drop zero-positive rows and update CSV export logic in age-bin evaluation
2026-01-16 17:51:00 +08:00
4068310a12
Refactor aggregation logic in age-bin results to handle pandas version compatibility
2026-01-16 17:24:53 +08:00
810c75e6d1
Add binary metrics computation and refactor evaluation logic in age-bin evaluation
2026-01-16 17:19:27 +08:00
b1647d1b74
Refactor tqdm import handling and improve context sampling in utils.py
2026-01-16 16:57:35 +08:00
e47a7ce4d6
Add multi-GPU support for age-bin evaluation and refactor aggregation logic
2026-01-16 16:27:02 +08:00
7a1210b5b0
Fix tqdm import handling by removing try-except block and ensuring proper import
2026-01-16 16:14:16 +08:00
90dffc3211
Add evaluation scripts for age-bin time-dependent metrics and remove obsolete evaluation_time_dependent.py
2026-01-16 16:13:31 +08:00
502ddd153b
Refactor context index selection in evaluate_time_dependent to improve horizon-specific eligibility handling
2026-01-16 15:01:35 +08:00
34d8d8ce9d
Add evaluation and utility functions for time-dependent metrics
...
- Introduced `evaluate.py` for time-dependent evaluation of models, including data loading and model inference.
- Added `evaluation_time_dependent.py` to compute various evaluation metrics such as AUC, average precision, and precision/recall at specified thresholds.
- Implemented CIF calculation methods in `losses.py` for different loss types, including exponential and piecewise exponential models.
- Created utility functions in `utils.py` for context selection and multi-hot encoding of events within specified horizons.
2026-01-16 14:55:09 +08:00
660dc969bc
Remove models_eval_example.json file containing model evaluation configurations.
2026-01-16 13:09:38 +08:00
2f46acf2bd
Add Piecewise Exponential CIF Loss and update model evaluation for PWE
2026-01-15 11:36:24 +08:00
d8b322cbee
Enhance Event Rate@K and Recall@K computations with random ranking baseline and additional metrics
2026-01-11 00:52:35 +08:00
4d53f52aa1
Refactor AUC computation methods and introduce Event Rate@K for cross-cause prioritization
2026-01-11 00:47:56 +08:00
d87752d1f8
Enhance evaluation metrics and progress visualization in model evaluation
...
- Introduced parallel processing for per-disease metrics using ThreadPoolExecutor.
- Added command-line arguments for metric workers and progress visualization options.
- Refactored evaluation functions to compute metrics for all diseases and summarize results.
- Updated output CSV filenames for clarity and consistency.
2026-01-10 23:49:37 +08:00
87baef3ecf
Fix device transfer for sexes in predict_cifs_for_model function
2026-01-10 17:02:28 +08:00
029f147ab5
Implement feature X to enhance user experience and fix bug Y in module Z
2026-01-10 17:00:16 +08:00
b88518a03f
Refactor model configurations in models_eval_example.json: update names, add loss_type and full_cov attributes for Delphi and SAP Delphi models.
2026-01-10 14:29:15 +08:00
320597dcde
Fix checkpoint path for sap_delphi_discrete_time_cif_partcov model in evaluation examples
2026-01-10 12:31:47 +08:00
b64ba101c2
Add sap_delphi_discrete_time_cif_partcov model configuration to evaluation examples
2026-01-10 12:22:14 +08:00
f795aa5604
Fix survival calculation in cifs_from_exponential_logits: broadcast mask for compatibility with tensor shapes
2026-01-10 11:42:03 +08:00
f231a2e4e5
Add model evaluation configurations for Delphi and SAP Delphi models
...
- Introduced a new JSON file `models_eval_example.json` containing evaluation configurations for various models.
- Included configurations for `delphi_fork` and `sap_delphi` models with different loss types and full covariance settings.
- Each model entry specifies the model name, type, loss type, full covariance flag, and checkpoint path.
2026-01-10 11:37:12 +08:00
3f15301f26
Enhance multi-GPU experiment runner: add log directory option, improve logging with sanitized filenames, and capture command output in log files.
2026-01-09 18:46:59 +08:00
5f8360a24a
Update experiments configuration: replace PiecewiseExponentialLoss with DiscreteTimeCIFNLLLoss for delphi_fork and sap_delphi models
2026-01-09 18:44:16 +08:00
739eb326f2
Fix formatting in DiscreteTimeCIFNLLLoss documentation for clarity
2026-01-09 18:34:01 +08:00
209dde2299
Refactor loss functions and model architecture: replace PiecewiseExponentialLoss with DiscreteTimeCIFNLLLoss, update Trainer to use SimpleHead, and modify argument parsing for new loss type.
2026-01-09 18:31:38 +08:00
880fd53a4b
Enhance Trainer class: add logging for WeibullNLLLoss parameters during training and validation
2026-01-09 13:48:36 +08:00
8723bf7600
Enhance Trainer class: add delta_scale logging for improved training statistics
2026-01-09 13:28:11 +08:00
dc34d51864
Add rank parameter to TrainConfig and update argument parsing for low-rank parameterization
2026-01-09 13:18:09 +08:00
