overwrite=None is deprecated. Use force_recompute instead.",
+ "overwrite=None is deprecated. Use force_recompute instead.",
DeprecationWarning,
)
if sample_size > 1 or sample_size <= 0:
@@ -1529,20 +1531,19 @@ def _return_correlation(
" `force_recompute=True` to override."
)
return getattr(self, "_" + "_".join(method.split()))
+ elif method == "pearson":
+ self._calc_pearson(corr_df, continuous_columns)
+ return self._pearson
+ elif method == "cramers v":
+ self._calc_cramers_v(corr_df, categorical_columns)
+ return self._cramers_v
+ elif method == "correlation ratio":
+ self._calc_correlation_ratio(
+ corr_df, categorical_columns, continuous_columns
+ )
+ return self._correlation_ratio
else:
- if method == "pearson":
- self._calc_pearson(corr_df, continuous_columns)
- return self._pearson
- elif method == "cramers v":
- self._calc_cramers_v(corr_df, categorical_columns)
- return self._cramers_v
- elif method == "correlation ratio":
- self._calc_correlation_ratio(
- corr_df, categorical_columns, continuous_columns
- )
- return self._correlation_ratio
- else:
- raise ValueError(f"The {method} method is not supported.")
+ raise ValueError(f"The {method} method is not supported.")
@runtime_dependency(module="IPython", install_from=OptionalDependency.NOTEBOOK)
def _reduce_dim_for_wide_dataset(
@@ -1551,8 +1552,11 @@ def _reduce_dim_for_wide_dataset(
min_cores_for_correlation = 2
n_rows, n_columns = self.shape
- from IPython.core.display import display, HTML
+ from IPython.display import HTML
+
+ from ads.common.utils import get_display
+ display = get_display()
if utils.get_cpu_count() <= min_cores_for_correlation:
msg = (
f"Not attempting to calculate correlations, too few cores ({utils.get_cpu_count()}) "
@@ -1695,14 +1699,12 @@ def show_corr(
if correlation_target:
if correlation_target not in features_list:
- raise ValueError(
- "correlation_target has to be in {}.".format(features_list)
- )
+ raise ValueError(f"correlation_target has to be in {features_list}.")
force_recompute = deprecate_variable(
overwrite,
force_recompute,
- f"overwrite=None is deprecated. Use force_recompute instead.",
+ "overwrite=None is deprecated. Use force_recompute instead.",
DeprecationWarning,
)
@@ -1787,7 +1789,7 @@ def show_in_notebook(
html_summary += "%s" % self.description html_summary += "
Note: Visualizations use a sampled subset of the dataset, this is to improve plotting performance. The sample size is calculated to be statistically - significant within the confidence level: {} and confidence interval: {}. + significant within the confidence level: {DatasetDefaults.sampling_confidence_level} and confidence interval: {DatasetDefaults.sampling_confidence_interval}. - The sampled data has {:,} rows + The sampled data has {sub_samp_df.shape[0]:,} rows
@@ -1818,11 +1820,7 @@ def show_in_notebook( - """.format( - DatasetDefaults.sampling_confidence_level, - DatasetDefaults.sampling_confidence_interval, - sub_samp_df.shape[0], - ) + """ html_summary += "" @@ -1892,7 +1890,7 @@ def show_in_notebook( force_recompute = deprecate_variable( overwrite, force_recompute, - f"overwrite=None is deprecated. Use force_recompute instead.",
+ "overwrite=None is deprecated. Use force_recompute instead.",
DeprecationWarning,
)
plot_type = kwargs.pop("plot_type", "heatmap")
@@ -1909,8 +1907,9 @@ def show_in_notebook(
**kwargs,
)
- from IPython.core.display import display
+ from ads.common.utils import get_display
+ display = get_display()
display(accordion)
# generate html for feature_distribution & warnings
diff --git a/ads/dataset/factory.py b/ads/dataset/factory.py
index c7bfb5139..7d368895e 100644
--- a/ads/dataset/factory.py
+++ b/ads/dataset/factory.py
@@ -1,58 +1,57 @@
#!/usr/bin/env python
-# -*- coding: utf-8; -*-
-# Copyright (c) 2020, 2024 Oracle and/or its affiliates.
+# Copyright (c) 2020, 2025 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
+import datetime
+import inspect
import os
import re
import warnings
+from typing import Callable, Tuple
+
+import fsspec
import oci
-import datetime
import pandas as pd
from fsspec.utils import infer_storage_options
-import inspect
-import fsspec
+from ocifs import OCIFileSystem
from ads.common import utils
+from ads.common.decorator.deprecate import deprecated
+from ads.common.decorator.runtime_dependency import (
+ OptionalDependency,
+ runtime_dependency,
+)
from ads.common.utils import is_same_class
from ads.dataset import logger
from ads.dataset.classification_dataset import (
BinaryClassificationDataset,
- MultiClassClassificationDataset,
BinaryTextClassificationDataset,
+ MultiClassClassificationDataset,
MultiClassTextClassificationDataset,
)
from ads.dataset.dataset import ADSDataset
from ads.dataset.forecasting_dataset import ForecastingDataset
from ads.dataset.helper import (
- get_feature_type,
- is_text_data,
- generate_sample,
DatasetDefaults,
- ElaboratedPath,
DatasetLoadException,
+ ElaboratedPath,
+ generate_sample,
+ get_feature_type,
+ is_text_data,
)
from ads.dataset.regression_dataset import RegressionDataset
from ads.type_discovery.type_discovery_driver import TypeDiscoveryDriver
from ads.type_discovery.typed_feature import (
+ CategoricalTypedFeature,
ContinuousTypedFeature,
DateTimeTypedFeature,
- CategoricalTypedFeature,
- OrdinalTypedFeature,
- GISTypedFeature,
DocumentTypedFeature,
+ GISTypedFeature,
+ OrdinalTypedFeature,
+ TypedFeature,
)
-from ads.type_discovery.typed_feature import TypedFeature
-from typing import Callable, Tuple
-from ocifs import OCIFileSystem
-from ads.common.decorator.runtime_dependency import (
- runtime_dependency,
- OptionalDependency,
-)
-from ads.common.decorator.deprecate import deprecated
default_snapshots_dir = None
default_storage_options = None
@@ -361,14 +360,11 @@ def list_snapshots(snapshot_dir=None, name="", storage_options=None, **kwargs):
# display in HTML format if sdk is run in notebook mode
if utils.is_notebook():
- from IPython.core.display import display
+ from ads.common.utils import get_display
+ display = get_display()
display(
- HTML(
- list_df.style.set_table_attributes("class=table")
- .hide()
- .to_html()
- )
+ HTML(list_df.style.set_table_attributes("class=table").hide().to_html())
)
return list_df
@@ -432,7 +428,7 @@ def _download_files(remote_files, local_path, overwrite=False):
os.makedirs(os.path.dirname(local_filepath), exist_ok=True)
with open(local_filepath, "wb") as f2:
f2.write(f1.read())
- except oci.exceptions.ServiceError as e:
+ except oci.exceptions.ServiceError:
raise FileNotFoundError(f"Unable to open file: {remote_file.path}")
return display_error, error_msg
@@ -600,7 +596,7 @@ def _get_dataset(
"It is not recommended to use an empty column as the target variable."
)
raise ValueError(
- f"We do not support using empty columns as the chosen target"
+ "We do not support using empty columns as the chosen target"
)
if is_same_class(target_type, ContinuousTypedFeature):
return RegressionDataset(
@@ -670,11 +666,7 @@ def _get_dataset(
is_same_class(target, DocumentTypedFeature)
or "text" in target_type["type"]
or "text" in target
- ):
- raise ValueError(
- f"The column {target} cannot be used as the target column."
- )
- elif (
+ ) or (
is_same_class(target_type, GISTypedFeature)
or "coord" in target_type["type"]
or "coord" in target
@@ -711,15 +703,15 @@ def read_tsv(path: str, **kwargs) -> pd.DataFrame:
def read_json(path: str, **kwargs) -> pd.DataFrame:
try:
return pd.read_json(path, **kwargs)
- except ValueError as e:
+ except ValueError:
return pd.read_json(
path, **utils.inject_and_copy_kwargs(kwargs, **{"lines": True})
)
@staticmethod
def read_libsvm(path: str, **kwargs) -> pd.DataFrame:
- from sklearn.datasets import load_svmlight_file
from joblib import Memory
+ from sklearn.datasets import load_svmlight_file
mem = Memory("./mycache")
@@ -808,7 +800,7 @@ def read_sql(cls, path: str, table: str = None, **kwargs) -> pd.DataFrame:
@staticmethod
def read_log(path, **kwargs):
- from ads.dataset.helper import parse_apache_log_str, parse_apache_log_datetime
+ from ads.dataset.helper import parse_apache_log_datetime, parse_apache_log_str
df = pd.read_csv(
path,
@@ -851,10 +843,11 @@ def read_html(path, html_table_index: int = None, **kwargs):
@staticmethod
@runtime_dependency(module="scipy", install_from=OptionalDependency.VIZ)
def read_arff(path, **kwargs):
- from scipy.io import arff
- import requests
from io import BytesIO, TextIOWrapper
+ import requests
+ from scipy.io import arff
+
data = None
if os.path.isfile(path):
data, _ = arff.loadarff(path)
@@ -881,7 +874,7 @@ def read_xml(path: str, **kwargs) -> pd.DataFrame:
-------
dataframe : pandas.DataFrame
"""
- import xml.etree.cElementTree as et
+ import xml.etree.ElementTree as et
def get_children(df, node, parent, i):
for name in node.attrib.keys():
@@ -969,18 +962,18 @@ def load_dataset(path: ElaboratedPath, reader_fn: Callable, **kwargs) -> pd.Data
dfs.append(data)
if len(dfs) == 0:
raise ValueError(
- f"We were unable to load the specified dataset. Read more here: "
- f"https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
- f"-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
+ "We were unable to load the specified dataset. Read more here: "
+ "https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
+ "-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
)
df = pd.concat(dfs)
if df is None:
raise ValueError(
- f"We were unable to load the specified dataset. Read more here: "
- f"https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
- f"-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
+ "We were unable to load the specified dataset. Read more here: "
+ "https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
+ "-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
)
if df.empty:
raise DatasetLoadException("Empty DataFrame, not producing a ADSDataset")
diff --git a/ads/dataset/helper.py b/ads/dataset/helper.py
index 777a6eb39..587c73d4a 100644
--- a/ads/dataset/helper.py
+++ b/ads/dataset/helper.py
@@ -1,55 +1,52 @@
#!/usr/bin/env python
-# -*- coding: utf-8; -*-
-# Copyright (c) 2020, 2023 Oracle and/or its affiliates.
+# Copyright (c) 2020, 2025 Oracle and/or its affiliates.
# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
import ast
import base64
import html
+import importlib
+import inspect
import io
import math
import os
-import warnings
import re
+import warnings
from collections import defaultdict
-import inspect
-import importlib
-from typing import Callable, List, Tuple, Union
-import fsspec
# from pandas.io.common import _compression_to_extension
-
from numbers import Number
+from typing import Callable, List, Tuple, Union
from urllib.parse import urlparse
+import fsspec
import numpy as np
import pandas as pd
-
from pandas.core.dtypes.common import (
- is_numeric_dtype,
is_bool_dtype,
is_categorical_dtype,
is_datetime64_any_dtype,
is_float_dtype,
+ is_numeric_dtype,
)
+from ads.common import utils
from ads.common.decorator.runtime_dependency import (
- runtime_dependency,
OptionalDependency,
+ runtime_dependency,
)
-from ads.common import utils
from ads.dataset import logger
from ads.type_discovery.type_discovery_driver import TypeDiscoveryDriver
from ads.type_discovery.typed_feature import (
+ CategoricalTypedFeature,
ContinuousTypedFeature,
DateTimeTypedFeature,
- CategoricalTypedFeature,
+ DocumentTypedFeature,
GISTypedFeature,
+ OrdinalTypedFeature,
TypedFeature,
UnknownTypedFeature,
- OrdinalTypedFeature,
- DocumentTypedFeature,
)
@@ -451,20 +448,19 @@ def generate_sample(
sample_size = calculate_sample_size(
n, min_size_to_sample, confidence_level, confidence_interval
)
+ elif min_size_to_sample < requested_sample_size < n:
+ logger.info(
+ f"Downsampling from {n} rows, to the user specified {requested_sample_size} rows for graphing."
+ )
+ sample_size = requested_sample_size
+ elif requested_sample_size >= n:
+ logger.info(f"Using the entire dataset of {n} rows for graphing.")
+ sample_size = n
else:
- if min_size_to_sample < requested_sample_size < n:
- logger.info(
- f"Downsampling from {n} rows, to the user specified {requested_sample_size} rows for graphing."
- )
- sample_size = requested_sample_size
- elif requested_sample_size >= n:
- logger.info(f"Using the entire dataset of {n} rows for graphing.")
- sample_size = n
- else:
- sample_size = min_size_to_sample
- logger.info(
- f"Downsampling from {n} rows, to {sample_size} rows for graphing."
- )
+ sample_size = min_size_to_sample
+ logger.info(
+ f"Downsampling from {n} rows, to {sample_size} rows for graphing."
+ )
if sample_size and len(df) > sample_size:
frac = min(1.0, sample_size * 1.05 / n)
@@ -581,14 +577,10 @@ def visualize_transformation(transformer_pipeline, text=None):
def format_label(stage):
if "FunctionTransformer" in str(transformer_pipeline.steps[stage][1].__class__):
- return "< {} >".format(
- html.escape(transformer_pipeline.steps[stage][1].func.__name__)
- )
+ return f"< {html.escape(transformer_pipeline.steps[stage][1].func.__name__)} >"
else:
is_ads = "ads" in str(transformer_pipeline.steps[stage][1].__class__)
- return "< {} >".format(
- transformer_pipeline.steps[stage][1].__class__.__name__
- )
+ return f"< {transformer_pipeline.steps[stage][1].__class__.__name__} >"
edges = [x[0] for x in transformer_pipeline.steps]
for i, edge in enumerate(list(zip(edges[:-1], edges[1:]))):
@@ -600,8 +592,11 @@ def format_label(stage):
graph = graphviz.Source(dot)
- from IPython.core.display import display, SVG
+ from IPython.display import SVG
+
+ from ads.common.utils import get_display
+ display = get_display()
display(SVG(graph.pipe(format="svg")))
@@ -700,7 +695,7 @@ def _get_imblearn_sampler(X, y):
k_neighbors = min(min_sample_size - 1, 5)
if k_neighbors == 0:
logger.warning(
- f"""k_neighbors is 0 as in the target there exists a class label that appeared only once.
+ """k_neighbors is 0 as in the target there exists a class label that appeared only once.
SMOTE will fail. Default to RandomOverSampler.
"""
)
@@ -806,9 +801,10 @@ def parse_apache_log_datetime(x):
Due to problems parsing the timezone (`%z`) with `datetime.strptime`, the
timezone will be obtained using the `pytz` library.
"""
- import pytz
from datetime import datetime
+ import pytz
+
dt = datetime.strptime(x[1:-7], "%d/%b/%Y:%H:%M:%S")
dt_tz = int(x[-6:-3]) * 60 + int(x[-3:-1])
return dt.replace(tzinfo=pytz.FixedOffset(dt_tz))
@@ -876,7 +872,7 @@ def get_dataset(
logger.warning(
"It is not recommended to use an empty column as the target variable."
)
- raise ValueError(f"We do not support using empty columns as the chosen target")
+ raise ValueError("We do not support using empty columns as the chosen target")
if utils.is_same_class(target_type, ContinuousTypedFeature):
return RegressionDataset(
df=df,
@@ -945,9 +941,7 @@ def get_dataset(
utils.is_same_class(target, DocumentTypedFeature)
or "text" in target_type["type"]
or "text" in target
- ):
- raise ValueError(f"The column {target} cannot be used as the target column.")
- elif (
+ ) or (
utils.is_same_class(target_type, GISTypedFeature)
or "coord" in target_type["type"]
or "coord" in target
@@ -1174,15 +1168,15 @@ def read_tsv(path: str, **kwargs) -> pd.DataFrame:
def read_json(path: str, **kwargs) -> pd.DataFrame:
try:
return pd.read_json(path, **kwargs)
- except ValueError as e:
+ except ValueError:
return pd.read_json(
path, **utils.inject_and_copy_kwargs(kwargs, **{"lines": True})
)
@staticmethod
def read_libsvm(path: str, **kwargs) -> pd.DataFrame:
- from sklearn.datasets import load_svmlight_file
from joblib import Memory
+ from sklearn.datasets import load_svmlight_file
mem = Memory("./mycache")
@@ -1271,7 +1265,7 @@ def read_sql(cls, path: str, table: str = None, **kwargs) -> pd.DataFrame:
@staticmethod
def read_log(path, **kwargs):
- from ads.dataset.helper import parse_apache_log_str, parse_apache_log_datetime
+ from ads.dataset.helper import parse_apache_log_datetime, parse_apache_log_str
df = pd.read_csv(
path,
@@ -1314,10 +1308,11 @@ def read_html(path, html_table_index: int = None, **kwargs):
@staticmethod
@runtime_dependency(module="scipy", install_from=OptionalDependency.VIZ)
def read_arff(path, **kwargs):
- from scipy.io import arff
- import requests
from io import BytesIO, TextIOWrapper
+ import requests
+ from scipy.io import arff
+
data = None
if os.path.isfile(path):
data, _ = arff.loadarff(path)
@@ -1344,7 +1339,7 @@ def read_xml(path: str, **kwargs) -> pd.DataFrame:
-------
dataframe : pandas.DataFrame
"""
- import xml.etree.cElementTree as et
+ import xml.etree.ElementTree as et
def get_children(df, node, parent, i):
for name in node.attrib.keys():
@@ -1432,18 +1427,18 @@ def load_dataset(path: ElaboratedPath, reader_fn: Callable, **kwargs) -> pd.Data
dfs.append(data)
if len(dfs) == 0:
raise ValueError(
- f"We were unable to load the specified dataset. Read more here: "
- f"https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
- f"-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
+ "We were unable to load the specified dataset. Read more here: "
+ "https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
+ "-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
)
df = pd.concat(dfs)
if df is None:
raise ValueError(
- f"We were unable to load the specified dataset. Read more here: "
- f"https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
- f"-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
+ "We were unable to load the specified dataset. Read more here: "
+ "https://docs.cloud.oracle.com/en-us/iaas/tools/ads"
+ "-sdk/latest/user_guide/loading_data/loading_data.html#specify-data-types-in-load-dataset"
)
if df.empty:
raise DatasetLoadException("Empty DataFrame, not producing a ADSDataset")
diff --git a/ads/dataset/plot.py b/ads/dataset/plot.py
index 9195ad34a..6bcf309ae 100644
--- a/ads/dataset/plot.py
+++ b/ads/dataset/plot.py
@@ -1,43 +1,39 @@
#!/usr/bin/env python
-# -*- coding: utf-8; -*-
-# Copyright (c) 2020, 2022 Oracle and/or its affiliates.
+# Copyright (c) 2020, 2025 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
import random
from collections import defaultdict
from math import pi
-import pandas as pd
-import numpy as np
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
+import pandas as pd
from matplotlib import colors as mcolors
-from ads.dataset.helper import _log_yscale_not_set
from ads.common.decorator.runtime_dependency import (
- runtime_dependency,
OptionalDependency,
+ runtime_dependency,
)
-from ads.common.utils import _log_plot_high_cardinality_warning, MAX_DISPLAY_VALUES
+from ads.common.utils import MAX_DISPLAY_VALUES, _log_plot_high_cardinality_warning
+from ads.dataset import logger
+from ads.dataset.helper import _log_yscale_not_set
from ads.type_discovery.latlon_detector import LatLonDetector
from ads.type_discovery.typed_feature import (
+ CategoricalTypedFeature,
+ ConstantTypedFeature,
ContinuousTypedFeature,
+ CreditCardTypedFeature,
DateTimeTypedFeature,
- ConstantTypedFeature,
DiscreteTypedFeature,
- CreditCardTypedFeature,
- ZipcodeTypedFeature,
- OrdinalTypedFeature,
- CategoricalTypedFeature,
GISTypedFeature,
+ OrdinalTypedFeature,
+ ZipcodeTypedFeature,
)
-from ads.dataset import logger
-
class Plotting:
def __init__(self, df, feature_types, x, y=None, plot_type="infer", yscale=None):
@@ -88,7 +84,7 @@ def select_best_plot(self):
for choice in choices:
if choice[1].__name__.lower().startswith(self.plot_type.lower()):
return choice
- logger.info("invalid plot_type: {}".format(self.plot_type))
+ logger.info(f"invalid plot_type: {self.plot_type}")
raise ValueError(
"plot_type: '%s' invalid, use one of: %s"
% (self.plot_type, ", ".join([x[0].__name__ for x in choices]))
@@ -97,7 +93,6 @@ def select_best_plot(self):
return choices[0]
def show_in_notebook(self, **kwargs):
-
"""
Visualizes the dataset by plotting the distribution of a feature or relationship between two features.
@@ -345,6 +340,7 @@ def _single_column_count_plot(x, data, yscale=None):
@runtime_dependency(module="folium", install_from=OptionalDependency.VIZ)
def _folium_map(x, data):
import folium.plugins
+
df = LatLonDetector.extract_x_y(data[x])
lat_min, lat_max, long_min, long_max = (
min(df.Y),
@@ -357,8 +353,9 @@ def _folium_map(x, data):
folium.plugins.HeatMap(df[["Y", "X"]]).add_to(m)
m.fit_bounds([[lat_min, long_min], [lat_max, long_max]])
- from IPython.core.display import display
+ from ads.common.utils import get_display
+ display = get_display()
display(m)
@staticmethod
@@ -372,7 +369,7 @@ def _multiple_pdf(x, y, data):
colors = dict(mcolors.BASE_COLORS, **mcolors.CSS4_COLORS)
hues = [
colors[x]
- for x in colors.keys()
+ for x in colors
if isinstance(colors[x], str) and colors[x].startswith("#")
]
@@ -385,7 +382,6 @@ def _multiple_pdf(x, y, data):
@runtime_dependency(module="seaborn", install_from=OptionalDependency.VIZ)
def _matplot(self, plot_method, figsize=(4, 3), **kwargs):
-
plt.style.use("seaborn-white")
plt.rc("xtick", labelsize="x-small")
@@ -398,13 +394,9 @@ def _matplot(self, plot_method, figsize=(4, 3), **kwargs):
#
# generate a title for the plot
#
- text = '{}, "{}" ({})'.format(
- plot_method.__name__.upper(), self.x, self.feature_types[self.x].type
- )
+ text = f'{plot_method.__name__.upper()}, "{self.x}" ({self.feature_types[self.x].type})'
if self.y:
- text = '{} vs "{}" ({})'.format(
- text, self.y, self.feature_types[self.y].type
- )
+ text = f'{text} vs "{self.y}" ({self.feature_types[self.y].type})'
plt.title(text, y=1.08)
plt.grid(linestyle="dotted")
@@ -425,7 +417,7 @@ def _matplot(self, plot_method, figsize=(4, 3), **kwargs):
# rename the y-axis label and x-axis label when "count" is the y-axis label
if self.y == "count":
- plt.xlabel("Column: {} values ".format(self.x))
+ plt.xlabel(f"Column: {self.x} values ")
plt.ylabel("instance count")
# add y-axis label as "count" when plot type is hist
@@ -454,7 +446,6 @@ def _generic_plot(self, plot_method, **kwargs):
@runtime_dependency(module="seaborn", install_from=OptionalDependency.VIZ)
def _get_plot_method(self):
-
#
# combos contains a dictionary with the key being a composite of the x and y types, the value will
# always be a list, possibly and empty list, indicating no match for combination
diff --git a/ads/dataset/recommendation.py b/ads/dataset/recommendation.py
index 4f3b45f7d..c4c4f1752 100644
--- a/ads/dataset/recommendation.py
+++ b/ads/dataset/recommendation.py
@@ -90,8 +90,9 @@ def _show_constant_fill_widget(self, column):
text.value = self.fill_nan_dict[column].value
self.fill_nan_dict[column] = text
- from IPython.core.display import display
+ from ads.common.utils import get_display
+ display = get_display()
display(text)
if self.control_buttons is not None:
# self.control_buttons.close()
@@ -149,8 +150,9 @@ def show_in_notebook(self):
@runtime_dependency(module="IPython", install_from=OptionalDependency.NOTEBOOK)
@runtime_dependency(module="ipywidgets", install_from=OptionalDependency.NOTEBOOK)
def _display(self):
- from IPython.core.display import display
+ from ads.common.utils import get_display
+ display = get_display()
if self.recommendation_type_index != len(self.recommendation_types):
if (
self.recommendation_types[self.recommendation_type_index]
diff --git a/ads/evaluations/evaluation_plot.py b/ads/evaluations/evaluation_plot.py
index fb89edaee..49ec16e95 100644
--- a/ads/evaluations/evaluation_plot.py
+++ b/ads/evaluations/evaluation_plot.py
@@ -1,26 +1,26 @@
#!/usr/bin/env python
-# -*- coding: utf-8; -*-
-# Copyright (c) 2020, 2023 Oracle and/or its affiliates.
+# Copyright (c) 2020, 2025 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
import base64
+import itertools
+import math
from io import BytesIO
+
import matplotlib as mpl
-import matplotlib.pyplot as plt
import matplotlib.lines as mlines
-from matplotlib.ticker import FormatStrFormatter
+import matplotlib.pyplot as plt
import numpy as np
-import math
+import pandas as pd
+from matplotlib.ticker import FormatStrFormatter
+
from ads.common import logger
from ads.common.decorator.runtime_dependency import (
- runtime_dependency,
OptionalDependency,
+ runtime_dependency,
)
-import itertools
-import pandas as pd
MAX_TITLE_LEN = 20
MAX_LEGEND_LEN = 10
@@ -36,7 +36,7 @@ def _fig_to_html(fig):
fig.savefig(tmpfile, format="png")
encoded = base64.b64encode(tmpfile.getvalue()).decode("utf-8")
- html = "
".format(encoded)
+ html = f"
"
return html
@@ -210,8 +210,11 @@ def render_legend_labels(label_dict):
pd.Series(label_dict, index=label_dict.keys()),
columns=["Shortened labels"],
)
- from IPython.core.display import display, HTML
+ from IPython.display import HTML
+
+ from ads.common.utils import get_display
+ display = get_display()
display(
HTML(
encodings.style.format(precision=4)
@@ -275,7 +278,7 @@ def _get_labels(classes, max_len=MAX_LEGEND_LEN):
conflict_dict = {}
for label in classes:
prefix = label if len(label) < max_len + 3 else label[:max_len] + "..."
- if conflict_dict.get(prefix, None) is None:
+ if conflict_dict.get(prefix) is None:
conflict_dict[prefix] = [label]
else:
conflict_dict[prefix].append(label)
@@ -344,12 +347,8 @@ def plot(
logger.info(
"Showing plot types: {}.".format(
", ".join(
- [
- "{}".format(EvaluationPlot._pretty_titles_map[str(p)])
- for p in plots
- ]
+ [f"{EvaluationPlot._pretty_titles_map[str(p)]}" for p in plots]
),
- ", ".join(["{}".format(x) for x in map(str, plots)]),
)
)
logger.info(plot_details)
@@ -424,7 +423,7 @@ def plot(
getattr(cls, "_" + plot_type)(ax, evaluation)
fig.tight_layout()
html_raw.append(_fig_to_html(fig))
- except KeyError as e:
+ except KeyError:
try:
if fig_title:
plt.close(fig=fig_title)
@@ -707,7 +706,7 @@ def _pretty_barh(
color=["teal", "blueviolet", "forestgreen", "peru", "y", "dodgerblue", "r"],
)
for j, v in enumerate(y):
- ax.annotate("{:.3f}".format(v), xy=(v / 2, j), va="center", ha="left")
+ ax.annotate(f"{v:.3f}", xy=(v / 2, j), va="center", ha="left")
if axis_labels:
if axis_labels[0]:
ax.set_xlabel(axis_labels[0], fontsize=12)
diff --git a/ads/evaluations/evaluator.py b/ads/evaluations/evaluator.py
index 694fe13e0..4da743402 100644
--- a/ads/evaluations/evaluator.py
+++ b/ads/evaluations/evaluator.py
@@ -1,19 +1,18 @@
#!/usr/bin/env python
-# -*- coding: utf-8; -*-
-# Copyright (c) 2020, 2023 Oracle and/or its affiliates.
+# Copyright (c) 2020, 2025 Oracle and/or its affiliates.
# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
-from cycler import cycler
import logging
+import re
+from typing import Any, List
+
import matplotlib as mpl
import numpy as np
-from numpy.typing import ArrayLike
import pandas as pd
-import re
+from cycler import cycler
+from numpy.typing import ArrayLike
from sklearn.preprocessing import LabelEncoder
-import tempfile
-from typing import List, Any
logging.getLogger("matplotlib").setLevel(logging.WARNING)
mpl.rcParams["image.cmap"] = "BuGn"
@@ -21,33 +20,32 @@
color=["teal", "blueviolet", "forestgreen", "peru", "y", "dodgerblue", "r"]
)
+from ads.common import logger
from ads.common.data import ADSData
from ads.common.decorator.runtime_dependency import (
- runtime_dependency,
OptionalDependency,
+ runtime_dependency,
)
-from ads.common.decorator.deprecate import deprecated
-from ads.common import logger
from ads.common.model import ADSModel
from ads.common.model_metadata import UseCaseType
from ads.dataset.dataset_with_target import ADSDatasetWithTarget
from ads.evaluations.evaluation_plot import EvaluationPlot
from ads.evaluations.statistical_metrics import (
- ModelEvaluator,
- DEFAULT_BIN_CLASS_METRICS,
- DEFAULT_MULTI_CLASS_METRICS,
- DEFAULT_REG_METRICS,
DEFAULT_BIN_CLASS_LABELS_MAP,
+ DEFAULT_BIN_CLASS_METRICS,
DEFAULT_MULTI_CLASS_LABELS_MAP,
+ DEFAULT_MULTI_CLASS_METRICS,
DEFAULT_REG_LABELS_MAP,
+ DEFAULT_REG_METRICS,
+ ModelEvaluator,
)
-from ads.model.generic_model import GenericModel, VERIFY_STATUS_NAME
+from ads.model.generic_model import VERIFY_STATUS_NAME, GenericModel
METRICS_TO_MINIMIZE = ["hamming_loss", "hinge_loss", "mse", "mae"]
POSITIVE_CLASS_NAMES = ["yes", "y", "t", "true", "1"]
-class Evaluator(object):
+class Evaluator:
"""
BETA FEATURE
Evaluator is the new and preferred way to evaluate a model of list of models.
@@ -388,8 +386,11 @@ def display(
>>> multi_evaluator.display(plots=["normalized_confusion_matrix",
... "precision_by_label", "recall_by_label", "f1_by_label"])
"""
- from IPython.core.display import display, HTML
+ from IPython.display import HTML
+
+ from ads.common.utils import get_display
+ display = get_display()
legend_labels = (
legend_labels if legend_labels is not None else self.legend_labels
)
@@ -458,9 +459,7 @@ def html(
"{html_metrics}
" ) @@ -584,7 +583,7 @@ def _pretty_label(df, labels, copy=True): return html_raw -class ADSEvaluator(object): +class ADSEvaluator: """ADS Evaluator class. This class holds field and methods for creating and using ADS evaluator objects. @@ -684,7 +683,7 @@ def __init__( """ if any(isinstance(m, ADSModel) for m in models): logger.warn( - f"ADSModel is being deprecated. Users should instead use GenericModel or one of its subclasses. More information here: https://accelerated-data-science.readthedocs.io/en/latest/user_guide/model_registration/introduction.html#register" + "ADSModel is being deprecated. Users should instead use GenericModel or one of its subclasses. More information here: https://accelerated-data-science.readthedocs.io/en/latest/user_guide/model_registration/introduction.html#register" ) self.evaluations = [] if isinstance(training_data, ADSDatasetWithTarget): @@ -814,8 +813,8 @@ def add_metrics(self, funcs, names): self.evaluations[0] = pd.concat([self.evaluations[0], pd_res]) if name not in self.metrics_to_show: self.metrics_to_show.append(name) - setattr(self, "train_evaluations", self.evaluations[0]) - setattr(self, "test_evaluations", self.evaluations[1]) + self.train_evaluations = self.evaluations[0] + self.test_evaluations = self.evaluations[1] def del_metrics(self, names): """Removes the listed metrics from the evaluator object it is called on. @@ -901,8 +900,8 @@ def add_models(self, models, show_full_name=False): ) self.evaluations = [total_train_metrics, total_test_metrics] - setattr(self, "train_evaluations", self.evaluations[0]) - setattr(self, "test_evaluations", self.evaluations[1]) + self.train_evaluations = self.evaluations[0] + self.test_evaluations = self.evaluations[1] def del_models(self, names): """Removes the listed models from the evaluator object it is called on. @@ -1044,7 +1043,7 @@ def calculate_cost( cost_df = pd.DataFrame({"model": list_of_model, "cost": cost_per_model}) return cost_df - class EvaluationMetrics(object): + class EvaluationMetrics: """Class holding evaluation metrics. Attributes @@ -1199,8 +1198,11 @@ def _display_metrics(df, data_name, labels, precision): ------- Nothing """ - from IPython.core.display import display, HTML + from IPython.display import HTML + + from ads.common.utils import get_display + display = get_display() display( HTML( _pretty_label(df, labels) diff --git a/ads/explanations/mlx_global_explainer.py b/ads/explanations/mlx_global_explainer.py index 22f14b39c..854ced9a8 100644 --- a/ads/explanations/mlx_global_explainer.py +++ b/ads/explanations/mlx_global_explainer.py @@ -1,26 +1,26 @@ #!/usr/bin/env python -# -*- coding: utf-8; -*- -# Copyright (c) 2020, 2023 Oracle and/or its affiliates. +# Copyright (c) 2020, 2025 Oracle and/or its affiliates. # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/ -import numpy as np from abc import ABC, abstractmethod +import numpy as np + from ads.common import logger, utils +from ads.common.decorator.deprecate import deprecated +from ads.common.decorator.runtime_dependency import ( + OptionalDependency, + runtime_dependency, +) from ads.explanations.base_explainer import GlobalExplainer -from ads.explanations.mlx_interface import check_tabular_or_text -from ads.explanations.mlx_interface import init_lime_explainer -from ads.explanations.mlx_interface import init_permutation_importance_explainer from ads.explanations.mlx_interface import ( - init_partial_dependence_explainer, + check_tabular_or_text, init_ale_explainer, + init_lime_explainer, + init_partial_dependence_explainer, + init_permutation_importance_explainer, ) -from ads.common.decorator.runtime_dependency import ( - runtime_dependency, - OptionalDependency, -) -from ads.common.decorator.deprecate import deprecated class MLXGlobalExplainer(GlobalExplainer): @@ -186,9 +186,7 @@ def compute_feature_importance( ] if scoring_metric not in allowed_metrics and scoring_metric is not None: raise Exception( - "Scoring Metric not supported for this type of problem: {}, for problem type {}, the availble supported metrics are {}".format( - scoring_metric, self.mode_, allowed_metrics - ) + f"Scoring Metric not supported for this type of problem: {scoring_metric}, for problem type {self.mode_}, the availble supported metrics are {allowed_metrics}" ) if balance and sampling is None: sampling = {"technique": "random"} @@ -423,8 +421,11 @@ def show_in_notebook(self): # pragma: no cover pdp_plot = self.compute_partial_dependence([pdp_plot_feature_name]) # plot2 = pdp_plot.show_in_notebook() - from IPython.core.display import display, HTML + from IPython.display import HTML + from ads.common.utils import get_display + + display = get_display() display(HTML(plot1.data)) # display(HTML(plot1.data + plot2.data)) @@ -482,9 +483,7 @@ def configure_feature_importance(self, **kwargs): for k, _ in kwargs.items(): if k not in avail_args: raise ValueError( - "Unexpected argument for the feature importance explainer: {}".format( - k - ) + f"Unexpected argument for the feature importance explainer: {k}" ) if kwargs.get("client", None) is not None: @@ -528,9 +527,7 @@ def configure_partial_dependence(self, **kwargs): for k, _ in kwargs.items(): if k not in ["client"]: raise ValueError( - "Unexpected argument for the partial dependence explainer: {}".format( - k - ) + f"Unexpected argument for the partial dependence explainer: {k}" ) if kwargs.get("client", None) is not None: raise ValueError( @@ -563,9 +560,7 @@ def configure_accumulated_local_effects(self, **kwargs): for k, _ in kwargs.items(): if k not in ["client"]: raise ValueError( - "Unexpected argument for the accumulated local effects explainer: {}".format( - k - ) + f"Unexpected argument for the accumulated local effects explainer: {k}" ) if kwargs.get("client", None) is not None: raise ValueError( @@ -696,7 +691,6 @@ def _init_accumulated_local_effects(self, **kwargs): class MLXFeatureDependenceExplanation(ABC): - __name__ = "MLXFeatureDependenceExplanation" def __init__(self, fd, fd_exp): diff --git a/ads/feature_engineering/adsimage/image.py b/ads/feature_engineering/adsimage/image.py index f8d59deb8..9e6405b4a 100644 --- a/ads/feature_engineering/adsimage/image.py +++ b/ads/feature_engineering/adsimage/image.py @@ -1,7 +1,6 @@ #!/usr/bin/env python -# -*- coding: utf-8; -*- -# Copyright (c) 2022 Oracle and/or its affiliates. +# Copyright (c) 2020, 2025 Oracle and/or its affiliates. # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/ """ @@ -15,7 +14,7 @@ Examples -------- >>> from ads.feature_engineering import ADSImage ->>> from IPython.core.display import display +>>> from IPython.display import display >>> img = ADSImage.open("1.jpg") >>> display(img) >>> img.save("oci://