Switch to orbitalml name

Let's start switching to orbitalml name, it's easier to move to `orbital` (if the name becomes available) from `orbitalml` than from `mustela` in terms of communication and confusion.

Author: amol-

.github/workflows/tests.yml

@@ -30,10 +30,10 @@ jobs:
       - name: Set up PostgreSQL
         uses: harmon758/postgresql-action@v1
         with:
-          postgresql db: mustelatestdb
-          postgresql user: mustelatestuser
-          postgresql password: mustelatestpassword
+          postgresql db: orbitalmltestdb
+          postgresql user: orbitalmltestuser
+          postgresql password: orbitalmltestpassword
 
       - name: Run Test Suite
         run: |
-          uv run pytest -v --tb=short --disable-warnings --maxfail=1 --cov=mustela
+          uv run pytest -v --tb=short --disable-warnings --maxfail=1 --cov=orbitalml

LICENSE.md

@@ -1,6 +1,6 @@
 # MIT License
 
-Copyright (c) 2024 Mustela authors
+Copyright (c) 2024 OrbitalML authors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.rst

@@ -1,4 +1,4 @@
-Mustela
+OrbitalML
 =======
 
 Convert SKLearn pipelines into SQL queries for execution in a database
@@ -14,15 +14,15 @@ See `examples` directory for example pipelines.
 **Note**::
 
     Not all transformations and models can be represented as SQL queries,
-    so Mustela might not be able to implement the specific pipeline you are using.
+    so OrbitalML might not be able to implement the specific pipeline you are using.
 
 Getting Started
 ----------------
 
-Install Mustela::
+Install OrbitalML::
 
-    $ git clone https://github.com/posit-dev/mustela.git
-    $ pip install ./mustela
+    $ git clone https://github.com/posit-dev/orbitalml.git
+    $ pip install ./orbitalml
 
 Prepare some data::
 
@@ -34,7 +34,7 @@ Prepare some data::
     iris = load_iris(as_frame=True)
     iris_x = iris.data.set_axis(COLUMNS, axis=1)
 
-    # SQL and Mustela don't like dots in column names, replace them with underscores
+    # SQL and OrbitalML don't like dots in column names, replace them with underscores
     iris_x.columns = COLUMNS = [cname.replace(".", "_") for cname in COLUMNS]
 
     X_train, X_test, y_train, y_test = train_test_split(
@@ -57,21 +57,21 @@ Define a Scikit-Learn pipeline and train it::
     )
     pipeline.fit(X_train, y_train)
 
-Convert the pipeline to Mustela::
+Convert the pipeline to OrbitalML::
 
-    import mustela
-    import mustela.types
+    import orbitalml
+    import orbitalml.types
 
-    mustela_pipeline = mustela.parse_pipeline(pipeline, features={
-        "sepal_length": mustela.types.DoubleColumnType(),
-        "sepal_width": mustela.types.DoubleColumnType(),
-        "petal_length": mustela.types.DoubleColumnType(),
-        "petal_width": mustela.types.DoubleColumnType(),
+    orbitalml_pipeline = orbitalml.parse_pipeline(pipeline, features={
+        "sepal_length": orbitalml.types.DoubleColumnType(),
+        "sepal_width": orbitalml.types.DoubleColumnType(),
+        "petal_length": orbitalml.types.DoubleColumnType(),
+        "petal_width": orbitalml.types.DoubleColumnType(),
     })
 
 You can print the pipeline to see the result::
 
-    >>> print(mustela_pipeline)
+    >>> print(orbitalml_pipeline)
 
     ParsedPipeline(
         features={
@@ -107,7 +107,7 @@ You can print the pipeline to see the result::
 
 Now we can generate the SQL from the pipeline::
 
-    sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+    sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
 
 And check the resulting query::
 
@@ -141,7 +141,7 @@ by running the scikitlearn pipeline on the same set of data::
 Supported Models
 -----------------
 
-Mustela currently supports the following models:
+OrbitalML currently supports the following models:
 
 - Linear Regression
 - Logistic Regression

examples/README.rst

@@ -1,2 +1,2 @@
 A few examples with significant test cases
-that show how to use the mustela library.
+that show how to use the orbitalml library.

examples/minimal.py

@@ -7,15 +7,15 @@
 
 import duckdb
 
-import mustela
-import mustela.types
+import orbitalml
+import orbitalml.types
 
 COLUMNS = ["sepal.length", "sepal.width", "petal.length", "petal.width"]
 
 iris = load_iris(as_frame=True)
 iris_x = iris.data.set_axis(COLUMNS, axis=1)
 
-# SQL and Mustela don't like dots in column names, replace them with underscores
+# SQL and OrbitalML don't like dots in column names, replace them with underscores
 iris_x.columns = COLUMNS = [cname.replace(".", "_") for cname in COLUMNS]
 
 X_train, X_test, y_train, y_test = train_test_split(
@@ -33,15 +33,15 @@
 pipeline.fit(X_train, y_train)
 
 
-mustela_pipeline = mustela.parse_pipeline(pipeline, features={
-    "sepal_length": mustela.types.DoubleColumnType(),
-    "sepal_width": mustela.types.DoubleColumnType(),
-    "petal_length": mustela.types.DoubleColumnType(),
-    "petal_width": mustela.types.DoubleColumnType(),
+orbitalml_pipeline = orbitalml.parse_pipeline(pipeline, features={
+    "sepal_length": orbitalml.types.DoubleColumnType(),
+    "sepal_width": orbitalml.types.DoubleColumnType(),
+    "petal_length": orbitalml.types.DoubleColumnType(),
+    "petal_width": orbitalml.types.DoubleColumnType(),
 })
-print(mustela_pipeline)
+print(orbitalml_pipeline)
 
-sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
 print("\nGenerated Query for DuckDB:")
 print(sql)
 print("\nPrediction with SQL")

examples/pipeline_boosted_tree_classifier.py

@@ -12,14 +12,14 @@
 from sklearn.pipeline import Pipeline
 from sklearn.preprocessing import OneHotEncoder, StandardScaler
 
-import mustela
-import mustela.types
+import orbitalml
+import orbitalml.types
 
 PRINT_SQL = int(os.environ.get("PRINTSQL", "0"))
 ASSERT = int(os.environ.get("ASSERT", "0"))
 
 logging.basicConfig(level=logging.INFO)
-logging.getLogger("mustela").setLevel(logging.INFO)  # Set DEBUG to see translation process.
+logging.getLogger("orbitalml").setLevel(logging.INFO)  # Set DEBUG to see translation process.
 
 # Load Ames Housing for classification
 ames = fetch_openml(name="house_prices", as_frame=True)
@@ -92,23 +92,23 @@ def categorize_price(price: float) -> str:
 
 model.fit(X, y)
 
-# Convert types from numpy to mustela types
-features = mustela.types.guess_datatypes(X)
+# Convert types from numpy to orbitalml types
+features = orbitalml.types.guess_datatypes(X)
 
 # Target only 5 rows, so that it's easier for a human to understand
 data_sample = X.head(5)
 
 # Convert the model to an execution pipeline
-mustela_pipeline = mustela.parse_pipeline(model, features=features)
-print(mustela_pipeline)
+orbitalml_pipeline = orbitalml.parse_pipeline(model, features=features)
+print(orbitalml_pipeline)
 
 # Translate the pipeline to a query
 ibis_table = ibis.memtable(data_sample, name="DATA_TABLE")
-ibis_expression = mustela.translate(ibis_table, mustela_pipeline)
+ibis_expression = orbitalml.translate(ibis_table, orbitalml_pipeline)
 
 con = ibis.duckdb.connect()
 if PRINT_SQL:
-    sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+    sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
     print("\nGenerated Query for DuckDB:")
     print(sql)
     print("\nPrediction with SQL")

examples/pipeline_boosted_tree_regressor.py

@@ -12,14 +12,14 @@
 from sklearn.pipeline import Pipeline
 from sklearn.preprocessing import OneHotEncoder, StandardScaler
 
-import mustela
-import mustela.types
+import orbitalml
+import orbitalml.types
 
 PRINT_SQL = int(os.environ.get("PRINTSQL", "0"))
 ASSERT = int(os.environ.get("ASSERT", "0"))
 
 logging.basicConfig(level=logging.INFO)
-logging.getLogger("mustela").setLevel(logging.INFO)  # Set DEBUG to see translation process.
+logging.getLogger("orbitalml").setLevel(logging.INFO)  # Set DEBUG to see translation process.
 
 ames = fetch_openml(name="house_prices", as_frame=True)
 ames = ames.frame
@@ -36,7 +36,7 @@
 ]
 categorical_features = ames.select_dtypes(include=["object", "category"]).columns
 
-# Mustela requires the input and outputs of an imputer to
+# OrbitalML requires the input and outputs of an imputer to
 # be of the same type, as SimpleImputer has to compute the mean
 # the result is always a float. Which makes sense.
 # Let's convert all numeric features to doubles so
@@ -87,19 +87,19 @@
 # It's easier to understand if it's small
 data_sample = X.head(5)
 
-features = mustela.types.guess_datatypes(X)
-mustela_pipeline = mustela.parse_pipeline(model, features=features)
-print(mustela_pipeline)
+features = orbitalml.types.guess_datatypes(X)
+orbitalml_pipeline = orbitalml.parse_pipeline(model, features=features)
+print(orbitalml_pipeline)
 
 ibis_table = ibis.memtable(data_sample, name="DATA_TABLE")
-ibis_expression = mustela.translate(ibis_table, mustela_pipeline)
+ibis_expression = orbitalml.translate(ibis_table, orbitalml_pipeline)
 con = ibis.duckdb.connect()
 
 if PRINT_SQL:
     con = ibis.duckdb.connect()
     print(con.compile(ibis_expression))
 
-    sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+    sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
     print("\nGenerated Query for DuckDB:")
     print(sql)
     print("\nPrediction with SQL")
@@ -114,7 +114,7 @@
 # NOTE: Interestingly the DuckDB optimizer has a bug on this query too
 #       and unless disabled the query never completes.
 #       That's why we run using SQLite.
-#       The Mustela optimizer when enabled is able to preoptimize the query
+#       The OrbitalML optimizer when enabled is able to preoptimize the query
 #       which seems to allow DuckDB to complete the query as probably the DuckDB
 #       optimizer has less work to do in that case.
 print("\nPrediction with Ibis")

examples/pipeline_decision_tree_classifier.py

@@ -12,14 +12,14 @@
 from sklearn.preprocessing import OneHotEncoder
 from sklearn.tree import DecisionTreeClassifier
 
-import mustela
-import mustela.types
+import orbitalml
+import orbitalml.types
 
 PRINT_SQL = int(os.environ.get("PRINTSQL", "0"))
 ASSERT = int(os.environ.get("ASSERT", "0"))
 
 logging.basicConfig(level=logging.INFO)
-logging.getLogger("mustela").setLevel(logging.INFO)  # Change to DEBUG to see each translation step.
+logging.getLogger("orbitalml").setLevel(logging.INFO)  # Change to DEBUG to see each translation step.
 
 iris = load_iris()
 df = pd.DataFrame(
@@ -82,11 +82,11 @@ def categorize_area(a: float) -> str:
 
 pipeline.fit(X, y)
 
-features = mustela.types.guess_datatypes(X)
-print("Mustela Features:", features)
+features = orbitalml.types.guess_datatypes(X)
+print("OrbitalML Features:", features)
 
-mustela_pipeline = mustela.parse_pipeline(pipeline, features=features)
-print(mustela_pipeline)
+orbitalml_pipeline = orbitalml.parse_pipeline(pipeline, features=features)
+print(orbitalml_pipeline)
 
 # Test data
 example_data = pa.table(
@@ -101,10 +101,10 @@ def categorize_area(a: float) -> str:
 
 con = ibis.duckdb.connect()
 ibis_table = ibis.memtable(example_data, name="DATA_TABLE")
-ibis_expression = mustela.translate(ibis_table, mustela_pipeline)
+ibis_expression = orbitalml.translate(ibis_table, orbitalml_pipeline)
 
 if PRINT_SQL:
-    sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+    sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
     print("\nGenerated Query for DuckDB:")
     print(sql)
     print("\nPrediction with SQL")

examples/pipeline_decision_tree_regressor.py

@@ -12,14 +12,14 @@
 from sklearn.preprocessing import OneHotEncoder
 from sklearn.tree import DecisionTreeRegressor
 
-import mustela
-import mustela.types
+import orbitalml
+import orbitalml.types
 
 PRINT_SQL = int(os.environ.get("PRINTSQL", "0"))
 ASSERT = int(os.environ.get("ASSERT", "0"))
 
 logging.basicConfig(level=logging.INFO)
-logging.getLogger("mustela").setLevel(logging.INFO)  # Set DEBUG to see translation process.
+logging.getLogger("orbitalml").setLevel(logging.INFO)  # Set DEBUG to see translation process.
 
 # Carica il dataset
 iris = load_iris()
@@ -63,13 +63,13 @@
 
 pipeline.fit(X, y)
 
-# Converti le feature per Mustela
-features = mustela.types.guess_datatypes(X)
-print("Mustela Features:", features)
+# Converti le feature per OrbitalML
+features = orbitalml.types.guess_datatypes(X)
+print("OrbitalML Features:", features)
 
-# Converti la pipeline in SQL con Mustela
-mustela_pipeline = mustela.parse_pipeline(pipeline, features=features)
-print(mustela_pipeline)
+# Converti la pipeline in SQL con OrbitalML
+orbitalml_pipeline = orbitalml.parse_pipeline(pipeline, features=features)
+print(orbitalml_pipeline)
 
 # Test data
 example_data = pa.table(
@@ -81,13 +81,13 @@
     }
 )
 
-# Genera la query SQL con Mustela
+# Genera la query SQL con OrbitalML
 ibis_table = ibis.memtable(example_data, name="DATA_TABLE")
-ibis_expression = mustela.translate(ibis_table, mustela_pipeline)
+ibis_expression = orbitalml.translate(ibis_table, orbitalml_pipeline)
 
 con = ibis.duckdb.connect()
 if PRINT_SQL:
-    sql = mustela.export_sql("DATA_TABLE", mustela_pipeline, dialect="duckdb")
+    sql = orbitalml.export_sql("DATA_TABLE", orbitalml_pipeline, dialect="duckdb")
     print("\nGenerated Query for DuckDB:")
     print(sql)
     print("\nPrediction with SQL")