rename functions, e.g. build-models to train-model

README.md

@@ -79,15 +79,15 @@ Run the pipeline to train the model used by the OpenPredict API.
 From a Python script:
 
 ```python
-from openpredict.openpredict_omim_drugbank import build_drug_disease_classifier
+from openpredict.openpredict_omim_drugbank import train_drug_disease_classifier
 
-build_drug_disease_classifier()
+train_drug_disease_classifier()
 ```
 
 Or using the command line:
 
 ```bash
-openpredict build-models
+openpredict train-model
 ```
 
 > Work in progress

docker-compose.yml

@@ -7,7 +7,7 @@ services:
     # Outdated image: docker.pkg.github.com/maastrichtu-ids/translator-openpredict/openpredict-api:latest
     # Outdated image: umids/openpredict-api:latest
     restart: unless-stopped
-    # command: build-models
+    # command: train-model
     command: start-api --server-url https://openpredict.137.120.31.102.nip.io
     ports:
       - 8808:8808

docs/README.md

@@ -16,7 +16,7 @@
   * [trainModel](#.openpredict.openpredict_omim_drugbank.trainModel)
   * [multimetric\_score](#.openpredict.openpredict_omim_drugbank.multimetric_score)
   * [evaluate](#.openpredict.openpredict_omim_drugbank.evaluate)
-  * [get\_drug\_disease\_classifier](#.openpredict.openpredict_omim_drugbank.build_drug_disease_classifier)
+  * [get\_drug\_disease\_classifier](#.openpredict.openpredict_omim_drugbank.train_drug_disease_classifier)
   * [query\_omim\_drugbank\_classifier](#.openpredict.openpredict_omim_drugbank.query_omim_drugbank_classifier)
 * [openpredict.openpredict\_api](#.openpredict.openpredict_api)
   * [start\_spark](#.openpredict.openpredict_api.start_spark)
@@ -267,11 +267,11 @@ Evaluate the trained classifier
 
 Scores
 
-<a name=".openpredict.openpredict_omim_drugbank.build_drug_disease_classifier"></a>
+<a name=".openpredict.openpredict_omim_drugbank.train_drug_disease_classifier"></a>
 #### get\_drug\_disease\_classifier
 
 ```python
-build_drug_disease_classifier()
+train_drug_disease_classifier()
 ```
 
 The main function to run the drug-disease similarities pipeline,

openpredict/__main__.py

@@ -3,7 +3,7 @@
 import sys
 
 from openpredict.openpredict_api import start_api as start_openpredict_api
-from openpredict.openpredict_omim_drugbank import build_drug_disease_classifier
+from openpredict.openpredict_omim_drugbank import train_drug_disease_classifier
 
 @click.command()
 @click.option(
@@ -19,8 +19,8 @@ def start_api(port, server_url, debug, start_spark):
 
 
 @click.command()
-def build_models():
-    build_drug_disease_classifier()
+def train_model():
+    train_drug_disease_classifier()
 
 
 @click.group()
@@ -29,7 +29,7 @@ def main(args=None):
     logging.basicConfig(stream=sys.stderr, level=logging.INFO)
 
 main.add_command(start_api)
-main.add_command(build_models)
+main.add_command(train_model)
 
 if __name__ == "__main__":
     sys.exit(main())

openpredict/openpredict_omim_drugbank.py

@@ -10,7 +10,7 @@
 from sklearn.model_selection import GroupKFold, StratifiedKFold
 from joblib import dump, load
 import pkg_resources
-from openpredict.build_utils import generate_classifier_metadata
+from openpredict.train_utils import generate_classifier_metadata
 
 def adjcencydict2matrix(df, name1, name2):
     """Convert dict to matrix
@@ -325,7 +325,7 @@ def evaluate(test_df, clf):
     return scores
 
 
-def build_drug_disease_classifier():
+def train_drug_disease_classifier():
     """The main function to run the drug-disease similarities pipeline, 
     and build the drug-disease classifier.
     It returns, and stores the generated classifier as a `.joblib` file 

openpredict/build_utils.py → openpredict/train_utils.py

@@ -24,7 +24,7 @@ def generate_classifier_metadata(classifier_id, scores, label="OpenPredict class
     clf_prop_prefix = OPENPREDICT_NAMESPACE + classifier_id + "/"
 
     if not (clf_uri, None, None) in g:
-        print('Generating RDF metadata for the classifier build at ' + TTL_METADATA_FILE)
+        print('Generating RDF metadata for the trained classifier at ' + TTL_METADATA_FILE)
         g.add((clf_uri, RDF.type, URIRef(MLS_NAMESPACE + 'ModelEvaluation')))
         g.add((clf_uri, RDFS.label, Literal(label)))
         g.add((clf_uri, URIRef('http://www.w3.org/ns/prov#generatedAtTime'), Literal(datetime.now(), datatype=XSD.dateTime)))

tests/run_function.py

@@ -1,14 +1,14 @@
 from joblib import load
 from openpredict.openpredict_api import get_predict
-from openpredict.openpredict_omim_drugbank import build_drug_disease_classifier
+from openpredict.openpredict_omim_drugbank import train_drug_disease_classifier
 from openpredict.build_utils import generate_classifier_metadata
 from datetime import datetime
 
 ## Run it:
 # python3 tests/run_function.py
 
 # time_start = datetime.now()
-# clf, scores = build_drug_disease_classifier()
+# clf, scores = train_drug_disease_classifier()
 
 # time_build = datetime.now()
 

tests/test_build_models.py → tests/test_train_model.py

@@ -1,9 +1,9 @@
 import pytest
-from openpredict.openpredict_omim_drugbank import build_drug_disease_classifier
+from openpredict.openpredict_omim_drugbank import train_drug_disease_classifier
 
-def test_build_drug_disease_classifier():
+def test_train_drug_disease_classifier():
     """Test the model to get drug-disease similarities (drugbank-omim)"""
-    clf, scores = build_drug_disease_classifier()
+    clf, scores = train_drug_disease_classifier()
 
     assert 0.80 < scores['precision'] < 0.95
     assert 0.60 < scores['recall'] < 0.80