Make numpy and sklearn optional requirements; include errors for functions requiring these packages #140

Author: smlindauer

setup.py

@@ -41,11 +41,13 @@ def get_file(filename):
     include_package_data=True,
     packages=find_packages(where="src"),
     package_dir={"": "src"},
-    python_requires=">=3.5",
-    install_requires=["pandas", "scikit-learn", "requests", "pyyaml", "packaging"],
+    python_requires=">=3.6",
+    install_requires=["pandas", "requests", "pyyaml", "packaging"],
     extras_require={
         "swat": ["swat"],
         "GitPython": ["GitPython"],
+        "numpy": ["numpy"],
+        "scikit-learn": ["scikit-learn"],
         "kerberos": [
             'kerberos ; platform_system != "Windows"',
             'winkerberos ; platform_system == "Windows"',
@@ -66,7 +68,6 @@ def get_file(filename):
         "Intended Audience :: Developers",
         "Programming Language :: Python",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.5",
         "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",

src/sasctl/pzmm/writeScoreCode.py

@@ -2,7 +2,6 @@
 # SPDX-License-Identifier: Apache-2.0
 
 from pathlib import Path
-import numpy as np
 import re
 from ..core import platform_version
 from .._services.model_repository import ModelRepository as modelRepo
@@ -66,8 +65,8 @@ def writeScoreCode(
             columns. The writeScoreCode function currently supports int(64), float(64),
             and string data types for scoring. Providing a list of dict objects signals
             that the model files are being created from an MLFlow model.
-        targetDF : DataFrame
-            The `DataFrame` object contains the training data for the target variable. Note that
+        targetDF : pandas Series
+            The `DataFrame Series` object contains the training data for the target variable. Note that
             for MLFlow models, this can be set as None.
         modelPrefix : string
             The variable for the model name that is used when naming model files.
@@ -492,7 +491,7 @@ def score{modelPrefix}({inputVarList}):
                         )
                     )
                     if threshPrediction is None:
-                        threshPrediction = np.mean(targetDF)
+                        threshPrediction = targetDF.mean()
                     cls.pyFile.write(
                         """\n
     if ({metric0} >= {threshold}):

src/sasctl/pzmm/write_json_files.py

@@ -4,11 +4,9 @@
 # %%
 from pathlib import Path
 import sys
-
 import getpass
 import json
 import pandas as pd
-from sklearn import metrics
 import math
 
 
@@ -476,11 +474,17 @@ def calculateFitStat(
         """
         # If numpy inputs are supplied, then it is assumed that numpy is installed in the environment
         try:
-            # noinspection PyPackageRequirements
             import numpy as np
         except ImportError:
             np = None
 
+        try:
+            from sklearn import metrics
+        except ImportError:
+            raise RuntimeError(
+                "The 'scikit-learn' package is required to use the calculateFitStat function."
+            )
+
         nullJSONPath = Path(__file__).resolve().parent / "null_dmcas_fitstat.json"
         nullJSONDict = self.readJSONFile(nullJSONPath)