postgresml · montanalow · Oct 24, 2022 · Oct 24, 2022
diff --git a/pgml-extension/Cargo.toml b/pgml-extension/Cargo.toml
@@ -57,3 +57,6 @@ panic = "unwind"
 opt-level = 3
 lto = "fat"
 codegen-units = 1
+
+[profile.test]
+opt-level = 3
diff --git a/pgml-extension/docker/entrypoint.sh b/pgml-extension/docker/entrypoint.sh
@@ -13,9 +13,7 @@ echo "Creating user and database..."
 (createdb -U postgres -h 127.0.0.1 pgml_development 2> /dev/null) || true
 
 echo "Installing pgml extension..."
-psql -U postgres -h 127.0.0.1 pgml_development -c 'CREATE EXTENSION IF NOT EXISTS pgml'
-
-psql -U postgres -h 127.0.0.1 pgml_development -f tests/test.sql -P pager
+psql -U postgres -h 127.0.0.1 pgml_development -f sql/setup_examples.sql -P pager
 
 echo "Ready!"
 if [[ ! -z $@ ]]; then

diff --git a/pgml-extension/examples/binary_classification.sql b/pgml-extension/examples/binary_classification.sql
@@ -1,5 +1,6 @@
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
 
 SELECT pgml.load_dataset('breast_cancer');
 
@@ -45,7 +46,8 @@ SELECT malignant, pgml.predict(
         "worst fractal dimension"
     ]
 ) AS prediction
-FROM pgml.breast_cancer 
+FROM pgml.breast_cancer
+ORDER BY random()
 LIMIT 10;
 
 -- view raw class probabilities
@@ -85,6 +87,7 @@ SELECT malignant, pgml.predict_proba(
     ]
 ) AS prediction
 FROM pgml.breast_cancer
+ORDER BY random()
 LIMIT 10;
 
 --

diff --git a/pgml-extension/examples/image_classification.sql b/pgml-extension/examples/image_classification.sql
@@ -11,6 +11,7 @@
 
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
 
 SELECT pgml.load_dataset('digits');
 

diff --git a/pgml-extension/examples/joint_regression.sql b/pgml-extension/examples/joint_regression.sql
@@ -4,6 +4,7 @@
 
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
 
 SELECT pgml.load_dataset('linnerud');
 

diff --git a/pgml-extension/examples/multi_classification.sql b/pgml-extension/examples/multi_classification.sql
@@ -1,5 +1,6 @@
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
 
 SELECT pgml.load_dataset('iris');
 

diff --git a/pgml-extension/examples/regression.sql b/pgml-extension/examples/regression.sql
@@ -11,7 +11,7 @@
 
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
-\timing
+\timing on
 
 SELECT pgml.load_dataset('diabetes');
 

diff --git a/pgml-extension/examples/transformers.sql b/pgml-extension/examples/transformers.sql
@@ -1,5 +1,7 @@
 -- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
+
 
 SELECT pgml.transform(
     'translation_en_to_fr',

diff --git a/pgml-extension/examples/vectors.sql b/pgml-extension/examples/vectors.sql
@@ -1,4 +1,6 @@
+-- Exit on error (psql)
 \set ON_ERROR_STOP true
+\timing on
 
 -- Elementwise arithmetic w/ constants
 SELECT pgml.add(ARRAY[1.0, 2.0, 3.0], 3);

diff --git a/pgml-extension/sql/setup_examples.sql b/pgml-extension/sql/setup_examples.sql
@@ -0,0 +1,28 @@
+---
+--- Create the extension and load sample data into a database.
+---
+--- Usage:
+---
+---   $ cargo pgx run --release
+---   $ psql -P pager-off -h localhost -p 28813 -d pgml -f sql/setup_examples.sql
+---
+\set ON_ERROR_STOP true
+\timing on
+
+-- The intention is to only allow setup_examples.sql to run on a database that
+-- has not had example data installed before, e.g. docker run. This should
+-- error and stop the process if the extension is already present.
+CREATE EXTENSION pgml;
+
+SELECT pgml.load_dataset('breast_cancer');
+SELECT pgml.load_dataset('diabetes');
+SELECT pgml.load_dataset('digits');
+SELECT pgml.load_dataset('iris');
+SELECT pgml.load_dataset('linnerud');
+SELECT pgml.load_dataset('wine');
+
+\i examples/binary_classification.sql
+\i examples/image_classification.sql
+\i examples/joint_regression.sql
+\i examples/multi_classification.sql
+\i examples/regression.sql
diff --git a/pgml-extension/tests/test.sql b/pgml-extension/tests/test.sql
@@ -1,9 +1,18 @@
 ---
 --- Test our functions.
 ---
+--- Usage:
+---
+---   $ cargo pgx run --release
+---   $ psql -P pager-off -h localhost -p 28813 -d pgml -f tests/test.sql
+---
 \set ON_ERROR_STOP true
+\timing on
 
-\timing
+-- Reset the test database to a clean state on each run.
+DROP EXTENSION IF EXISTS pgml CASCADE;
+DROP SCHEMA IF EXISTS pgml CASCADE;
+CREATE EXTENSION pgml;
 
 SELECT pgml.load_dataset('breast_cancer');
 SELECT pgml.load_dataset('diabetes');