Merge pull request #8 from oracle/2.6.x

Release version 2.6.1

Author: liudmylaru

README.md

@@ -123,13 +123,13 @@ Multiple extra dependencies can be installed together. For example:
 
 ## Contributing
 
-This project welcomes contributions from the community. Before submitting a pull request, please review our contribution guide.
+This project welcomes contributions from the community. Before submitting a pull request, please review our contribution guide [CONTRIBUTING.md](https://github.com/oracle/accelerated-data-science/blob/main/CONTRIBUTING.md).
 
-Find Getting Started instructions for developers in [README-development.md](./README-development.md)
+Find Getting Started instructions for developers in [README-development.md](https://github.com/oracle/accelerated-data-science/blob/main/README-development.md)
 
 ## Security
 
-Please consult the security guide for our responsible security vulnerability disclosure process.
+Please consult the security guide [SECURITY.md](https://github.com/oracle/accelerated-data-science/blob/main/SECURITY.md) for our responsible security vulnerability disclosure process.
 
 ## License
 

ads/__init__.py

@@ -10,14 +10,13 @@
 import sys
 
 import IPython
-import oci
 from IPython import get_ipython
 from IPython.core.error import UsageError
-
 import matplotlib.font_manager  # causes matplotlib to regenerate its fonts
 import json
 
 import ocifs
+import oci
 from ads.common.decorator.deprecate import deprecated
 from ads.feature_engineering.accessor.series_accessor import ADSSeriesAccessor
 from ads.feature_engineering.accessor.dataframe_accessor import ADSDataFrameAccessor

ads/ads_version.json

@@ -1,3 +1,3 @@
 {
-  "version": "2.5.10"
+  "version": "2.6.1"
 }

ads/automl/driver.py

@@ -4,30 +4,29 @@
 # Copyright (c) 2020, 2022 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
-from __future__ import print_function, absolute_import, division
+from __future__ import absolute_import, division, print_function
 
+import numpy as np
+import pandas as pd
+import scipy
 import copy
-
+from ads.common import logger, utils
 from ads.common.model import ADSModel
-from ads.common import logger
 from ads.dataset import helper
-from ads.dataset.dataset_with_target import ADSDatasetWithTarget
 from ads.dataset.classification_dataset import (
     BinaryClassificationDataset,
+    BinaryTextClassificationDataset,
     MultiClassClassificationDataset,
     MultiClassTextClassificationDataset,
-    BinaryTextClassificationDataset,
 )
-from ads.dataset.regression_dataset import RegressionDataset
-
+from ads.dataset.dataset_with_target import ADSDatasetWithTarget
 from ads.dataset.pipeline import TransformerPipeline
+from ads.dataset.regression_dataset import RegressionDataset
 from ads.type_discovery.type_discovery_driver import TypeDiscoveryDriver
 from ads.type_discovery.typed_feature import (
     ContinuousTypedFeature,
     DiscreteTypedFeature,
 )
-from ads.common import utils
-
 
 dataset_task_map = {
     BinaryClassificationDataset: utils.ml_task_types.BINARY_CLASSIFICATION,
@@ -114,7 +113,7 @@ def __init__(
         >>> olabs_automl = OracleAutoMLProvider()
         >>> model, baseline = AutoML(train, provider=olabs_automl).train()
         """
-        from ads.automl.provider import OracleAutoMLProvider, BaselineAutoMLProvider
+        from ads.automl.provider import BaselineAutoMLProvider, OracleAutoMLProvider
 
         if hasattr(training_data, "transformer_pipeline"):
             self.transformer_pipeline = training_data.transformer_pipeline
@@ -154,7 +153,23 @@ def __init__(
                 or utils._is_dask_series(training_data.y)
                 else training_data.y
             )
-        self.target_name = y.name
+
+        if isinstance(y, pd.DataFrame):
+            if len(y.columns) != 1:
+                raise ValueError("Data must be 1-dimensional.")
+            y = y.squeeze()
+        elif isinstance(y, np.ndarray):
+            y = pd.Series(y)
+        if y.name:
+            self.target_name = str(y.name)
+        else:
+            y.name = str(0)
+            self.target_name = str(0)
+
+        if isinstance(X, np.ndarray):
+            X = pd.DataFrame(X)
+        elif isinstance(X, scipy.sparse.csr.csr_matrix):
+            X = pd.DataFrame(X.todense())
         self.feature_names = X.columns.values
         self.client = client
         class_names = y.unique()
@@ -256,8 +271,8 @@ def train(self, **kwargs):
         avail_n_cores = utils.get_cpu_count()
 
         warn_params = [
-            (10 ** 5, 4, "VM.Standard.E2.4"),
-            (10 ** 6, 16, "VM.Standard.2.16"),
+            (10**5, 4, "VM.Standard.E2.4"),
+            (10**6, 16, "VM.Standard.2.16"),
         ]
 
         # train using automl and baseline

ads/automl/provider.py

@@ -304,7 +304,9 @@ def decide_estimator(self, **kwargs):
 
 # An installation of oracle labs automl is required only for this class
 class OracleAutoMLProvider(AutoMLProvider, ABC):
-    def __init__(self, n_jobs=-1, loglevel=None, logger_override=None):
+    def __init__(
+        self, n_jobs=-1, loglevel=None, logger_override=None, model_n_jobs: int = 1
+    ):
         """
         The Oracle AutoML Provider automatically provides a tuned ML pipeline that best models the given a training
         dataset and a prediction task at hand.
@@ -317,6 +319,9 @@ def __init__(self, n_jobs=-1, loglevel=None, logger_override=None):
         loglevel : int
             The verbosity of output for Oracle AutoML. Can be specified using the Python logging module
             (https://docs.python.org/3/library/logging.html#logging-levels).
+        model_n_jobs: (optional, int). Defaults to 1.
+            Specifies the model parallelism used by AutoML.
+            This will be passed to the underlying model it is training.
         """
         try:
             self.automl = __import__("automl")
@@ -327,24 +332,13 @@ def __init__(self, n_jobs=-1, loglevel=None, logger_override=None):
         super(OracleAutoMLProvider, self).__init__()
         if loglevel is None:
             loglevel = logging.DEBUG if ads.debug_mode else logging.ERROR
-        else:
-            loglevel = loglevel
-        if logger_override:
-            logr = logger_override
-        else:
-            logr = logging.getLogger(__name__)
-        if "AMD" in self.cpuinfo.get_cpu_info().get("brand", "UNKNOWN-BRAND"):
-            # Disable intra-model parallelism for LightGBM and XGBoost libraries
-            # which seem to be unstable currently on AMD shapes
-            self.automl.init(
-                engine="local",
-                engine_opts={"n_jobs": n_jobs, "model_n_jobs": 1},
-                loglevel=loglevel,
-            )
-        else:
-            self.automl.init(
-                engine="local", engine_opts={"n_jobs": n_jobs}, logger=logr
-            )
+
+        self.automl.init(
+            engine="local",
+            engine_opts={"n_jobs": n_jobs, "model_n_jobs": model_n_jobs},
+            logger=logger_override,
+            loglevel=loglevel,
+        )
 
     def __repr__(self):
         super(OracleAutoMLProvider, self).__repr__()
@@ -620,15 +614,16 @@ def _decide_estimator(self, **kwargs):
             )
             if not self.col_types:
                 if len(self.X_train.columns) == 1:
-                    self.col_types = ['text']
+                    self.col_types = ["text"]
                 elif len(self.X_train.columns) == 2:
-                    self.col_types = ['text', 'text']
+                    self.col_types = ["text", "text"]
                 else:
-                    raise ValueError("We detected a text classification problem. Pass " \
-                        "in `col_types = [<type of column1>, <type of column2>, ...]`." \
-                        " Valid types are: ['categorical', 'numerical', 'text', 'datetime'," \
+                    raise ValueError(
+                        "We detected a text classification problem. Pass "
+                        "in `col_types = [<type of column1>, <type of column2>, ...]`."
+                        " Valid types are: ['categorical', 'numerical', 'text', 'datetime',"
                         " 'timedelta']."
-                        )
+                    )
 
         elif self.ml_task_type == ml_task_types.REGRESSION:
             est = self.automl.Pipeline(

ads/bds/auth.py

@@ -119,6 +119,7 @@ def refresh_ticket(
     ...                    auth='KERBEROS',
     ...                    kerberos_service_name="hive").cursor()
     """
+    keytab_path = os.path.abspath(os.path.expanduser(keytab_path))
     os.environ[KRB5_CONFIG] = os.path.abspath(os.path.expanduser(kerb5_path))
     if not os.path.exists(os.environ[KRB5_CONFIG]):
         raise FileNotFoundError(f"krb5 config file not found in {kerb5_path}.")