change init seed to prior seed

.github/workflows/ci-workflow.yml

@@ -29,7 +29,7 @@ jobs:
         ruff check .
     - name: Render makita templates
       run: |
-        asreview makita template basic -p basic -d .github/workflows/test_data/ --classifier nb --feature_extractor tfidf --query_strategy max --n_runs 1 --init_seed 1 --model_seed 2 --skip_wordclouds --overwrite --instances_per_query 2 --stop_if min --balance_strategy double | tee output_basic.txt
+        asreview makita template basic -p basic -d .github/workflows/test_data/ --classifier nb --feature_extractor tfidf --query_strategy max --n_runs 1 --prior_seed 1 --model_seed 2 --skip_wordclouds --overwrite --instances_per_query 2 --stop_if min --balance_strategy double | tee output_basic.txt
         grep -q "ERROR" output_basic.txt && exit 1 || true
         asreview makita template arfi -p arfi -d .github/workflows/test_data/ | tee output_arfi.txt
         grep -q "ERROR" output_arfi.txt && exit 1 || true

README.md

@@ -181,7 +181,7 @@ optional arguments:
   -p, --project_folder PROJECT_FOLDER       The folder the project will be rendered too     Default: The current working directory
   -d, --data_folder DATA_FOLDER             The dataset source folder                       Default: `Data` folder in working directory
   -j, --job_file JOB_FILE                   The name of the file with jobs                  Default: jobs.bat for Windows, otherwise jobs.sh.
-  --init_seed INIT_SEED                     Seed of the priors                              Seed is set to 535 by default.
+  --prior_seed prior_seed                   Seed of the priors                              Seed is set to 535 by default.
   --model_seed MODEL_SEED                   Seed of the models                              Seed is set to 165 by default.
   --template TEMPLATE                       Overwrite template with template file path 
   --platform PLATFORM                       Platform to run jobs: Windows, Darwin, Linux.   Default: the system of rendering templates.
@@ -213,7 +213,7 @@ optional arguments:
   -p, --project_folder PROJECT_FOLDER       The folder the project will be rendered too     Default: The current working directory
   -d, --data_folder DATA_FOLDER             The dataset source folder                       Default: `Data` folder in working directory
   -j, --job_file JOB_FILE                   The name of the file with jobs                  Default: jobs.bat for Windows, otherwise jobs.sh.
-  --init_seed INIT_SEED                     Seed of the priors.                             Seed is set to 535 by default.
+  --prior_seed prior_seed                   Seed of the priors.                             Seed is set to 535 by default.
   --model_seed MODEL_SEED                   Seed of the models.                             Seed is set to 165 by default.
   --template TEMPLATE                       Overwrite template with template file path.
   --platform PLATFORM                       Platform to run jobs: Windows, Darwin, Linux.   Default: the system of rendering templates.
@@ -243,7 +243,7 @@ optional arguments:
   -p, --project_folder PROJECT_FOLDER       The folder the project will be rendered too     Default: The current working directory
   -d, --data_folder DATA_FOLDER             The dataset source folder                       Default: `Data` folder in working directory
   -j, --job_file JOB_FILE                   The name of the file with jobs                  Default: jobs.bat for Windows, otherwise jobs.sh.
-  --init_seed INIT_SEED                     Seed of the priors.                             Seed is set to 535 by default.
+  --prior_seed prior_seed                   Seed of the priors.                             Seed is set to 535 by default.
   --model_seed MODEL_SEED                   Seed of the models.                             Seed is set to 165 by default.
   --template TEMPLATE                       Overwrite template with template file path.
   --platform PLATFORM                       Platform to run jobs: Windows, Darwin, Linux.   Default: the system of rendering templates.
@@ -300,7 +300,7 @@ optional arguments:
   -p, --project_folder PROJECT_FOLDER       The folder the project will be rendered too     Default: The current working directory
   -d, --data_folder DATA_FOLDER             The dataset source folder                       Default: `Data` folder in working directory
   -j, --job_file JOB_FILE                   The name of the file with jobs                  Default: jobs.bat for Windows, otherwise jobs.sh.
-  --init_seed INIT_SEED                     Seed of the priors.                             Seed is set to 535 by default.
+  --prior_seed prior_seed                   Seed of the priors.                             Seed is set to 535 by default.
   --model_seed MODEL_SEED                   Seed of the models.                             Seed is set to 165 by default.
   --template TEMPLATE                       Overwrite template with template file path.
   --platform PLATFORM                       Platform to run jobs: Windows, Darwin, Linux.   Default: the system of rendering templates.

asreviewcontrib/makita/entrypoint.py

@@ -53,7 +53,7 @@ def execute(self, argv):  # noqa: C901
             help="Set project folder path." "Default will use current directory.",
         )
         parser_template.add_argument(
-            "--init_seed",
+            "--prior_seed",
             type=int,
             default=535,
             help="Seed of the priors. 535 by default.",
@@ -304,7 +304,7 @@ def _get_template_args(self):
             "skip_wordclouds",
             "n_runs",
             "n_priors",
-            "init_seed",
+            "prior_seed",
             "model_seed",
             "classifier",
             "feature_extractor",

asreviewcontrib/makita/template_arfi.py

@@ -34,7 +34,7 @@ def get_dataset_specific_params(self, index, fp_dataset):
         n_priors = self.n_priors if self.n_priors is not None else 10
 
         priors = _get_priors(
-            fp_dataset, init_seed=self.init_seed + index, n_priors=n_priors
+            fp_dataset, prior_seed=self.prior_seed + index, n_priors=n_priors
         )
         return {
             "input_file": f"{fp_dataset.parent.name}/{fp_dataset.name}",
@@ -62,14 +62,14 @@ def get_template_specific_params(self, params):
             "balance_strategy": balance_strategy,
             "instances_per_query": self.instances_per_query,
             "stop_if": self.stop_if,
-            "init_seed": self.init_seed,
+            "prior_seed": self.prior_seed,
             "output_folder": self.paths.output_folder,
             "scripts_folder": self.paths.scripts_folder,
             "version": self.__version__,
         }
 
 
-def _get_priors(dataset, init_seed, n_priors):
+def _get_priors(dataset, prior_seed, n_priors):
     """Sample priors."""
 
     records = load_dataset(dataset, dataset_id=Path(dataset).name)
@@ -86,7 +86,7 @@ def _get_priors(dataset, init_seed, n_priors):
     if len(relevant_irrecord_ids) == 0:
         raise ValueError("No irrelevant records found.")
 
-    np.random.seed(init_seed)
+    np.random.seed(prior_seed)
 
     prior_irrelevant = list(
         np.random.choice(relevant_irrecord_ids, n_priors, replace=False)

asreviewcontrib/makita/template_base.py

@@ -20,7 +20,7 @@ def __init__(
         file_handler: FileHandler,
         paths: ProjectPaths,
         skip_wordclouds,
-        init_seed,
+        prior_seed,
         model_seed,
         balance_strategy,
         instances_per_query,
@@ -30,7 +30,7 @@ def __init__(
         self.datasets = datasets
         self.paths = paths
         self.skip_wordclouds = skip_wordclouds
-        self.init_seed = init_seed
+        self.prior_seed = prior_seed
         self.model_seed = model_seed
         self.balance_strategy = balance_strategy
         self.instances_per_query = instances_per_query

asreviewcontrib/makita/template_basic.py

@@ -29,7 +29,7 @@ def get_dataset_specific_params(self, index, fp_dataset):
             "input_file": f"{fp_dataset.parent.name}/{fp_dataset.name}",
             "input_file_stem": fp_dataset.stem,
             "model_seed": self.model_seed + index,
-            "init_seed": self.init_seed,
+            "prior_seed": self.prior_seed,
         }
 
     def get_template_specific_params(self, params):

asreviewcontrib/makita/template_multimodel.py

@@ -34,7 +34,7 @@ def get_dataset_specific_params(self, index, fp_dataset):
             "input_file": f"{fp_dataset.parent.name}/{fp_dataset.name}",
             "input_file_stem": fp_dataset.stem,
             "model_seed": self.model_seed + index,
-            "init_seed": self.init_seed,
+            "prior_seed": self.prior_seed,
         }
 
     def get_template_specific_params(self, params):

asreviewcontrib/makita/template_prior.py

@@ -175,7 +175,7 @@ def get_template_specific_params(self, params):
             "scripts_folder": self.paths.scripts_folder,
             "version": self.__version__,
             "model_seed": self.model_seed,
-            "init_seed": self.init_seed,
+            "prior_seed": self.prior_seed,
             "filepath_with_priors": 
             f"{filepath_with_priors.parent.name}/{filepath_with_priors.name}",
             "filepath_with_priors_stem": filepath_with_priors.stem,

asreviewcontrib/makita/templates/template_basic.txt.template

@@ -49,7 +49,7 @@ python -m asreview wordcloud {{ dataset.input_file }} -o {{ output_folder }}/fig
 # Simulate runs
 mkdir {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files
 {% for run in range(n_runs) %}
-python -m asreview simulate {{ dataset.input_file }} -s {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --init_seed {{ dataset.init_seed + run }} --seed {{ dataset.model_seed + run }} -m {{ classifier }} -e {{ feature_extractor }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
+python -m asreview simulate {{ dataset.input_file }} -o {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --prior-seed {{ dataset.prior_seed + run }} --seed {{ dataset.model_seed + run }} -m {{ classifier }} -e {{ feature_extractor }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
 python -m asreview metrics {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview -o {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/metrics/metrics_sim_{{ dataset.input_file_stem }}{{ "_{}".format(run) if n_runs > 1 else "" }}.json
 {% endfor %}
 

asreviewcontrib/makita/templates/template_multimodel.txt.template

@@ -55,7 +55,7 @@ mkdir {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files
 # Skipped {{ classifier }} + {{ feature_extraction }} + {{ query_strategy}} model
 {% else %}# Classifier = {{ classifier }}, Feature extractor = {{ feature_extraction }}, Query strategy = {{ query_strategy }}, Balance strategy = {{balance_strategy}}
 {% for run in range(n_runs) %}
-python -m asreview simulate {{ dataset.input_file }} -s {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}_{{ classifier }}_{{ feature_extraction }}_{{ query_strategy }}_{{ balance_strategy }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --model {{ classifier }} --query_strategy {{query_strategy}} --feature_extraction {{ feature_extraction }} --init_seed {{ dataset.init_seed + run }} --seed {{ dataset.model_seed }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
+python -m asreview simulate {{ dataset.input_file }} -s {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}_{{ classifier }}_{{ feature_extraction }}_{{ query_strategy }}_{{ balance_strategy }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --model {{ classifier }} --query_strategy {{query_strategy}} --feature_extraction {{ feature_extraction }} --prior_seed {{ dataset.prior_seed + run }} --seed {{ dataset.model_seed }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
 python -m asreview metrics {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/state_files/sim_{{ dataset.input_file_stem }}_{{ classifier }}_{{ feature_extraction }}_{{ query_strategy }}_{{ balance_strategy }}{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview -o {{ output_folder }}/simulation/{{ dataset.input_file_stem }}/metrics/metrics_sim_{{ dataset.input_file_stem }}_{{ classifier }}_{{ feature_extraction }}_{{ query_strategy }}_{{ balance_strategy }}{{ "_{}".format(run) if n_runs > 1 else "" }}.json
 {% endfor %}{% endif %}
 {% endfor %}

asreviewcontrib/makita/templates/template_prior.txt.template

@@ -42,7 +42,7 @@ python -m asreview wordcloud {{ filepath_without_priors }} -o {{ output_folder }
 python -m asreview simulate {{ filepath_with_priors }} -s {{ output_folder }}/simulation/state_files/sim_custom_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --seed {{ model_seed + run }} -m {{ classifier }} -e {{ feature_extractor }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }} --prior_idx {{ prior_idx }}
 python -m asreview metrics {{ output_folder }}/simulation/state_files/sim_custom_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview -o {{ output_folder }}/simulation/metrics/metrics_sim_custom_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.json
 
-python -m asreview simulate {{ filepath_without_priors }} -s {{ output_folder }}/simulation/state_files/sim_minimal_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --init_seed {{ init_seed + run }} --seed {{ model_seed + run }} -m {{ classifier }} -e {{ feature_extractor }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
+python -m asreview simulate {{ filepath_without_priors }} -s {{ output_folder }}/simulation/state_files/sim_minimal_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview --prior_seed {{ prior_seed + run }} --seed {{ model_seed + run }} -m {{ classifier }} -e {{ feature_extractor }} -q {{ query_strategy }} -b {{ balance_strategy }} --n_instances {{ instances_per_query }} --stop_if {{ stop_if }}
 python -m asreview metrics {{ output_folder }}/simulation/state_files/sim_minimal_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.asreview -o {{ output_folder }}/simulation/metrics/metrics_sim_minimal_priors{{ "_{}".format(run) if n_runs > 1 else "" }}.json
 
 {% endfor %}