Skip to content

Meta

Meta Examples

Examples of using the Meta class are listed at the bottom of this page Examples.

Meta: A class that provides high level information and summaries of Cloud Platform Artifacts. The Meta class provides 'account' information, configuration, etc. It also provides metadata for Artifacts, such as Data Sources, Feature Sets, Models, and Endpoints.

Meta

Bases: CloudMeta

Meta: A class that provides metadata functionality for Cloud Platform Artifacts.

Common Usage 
from workbench.api import Meta
meta = Meta()

# Get the AWS Account Info
meta.account()
meta.config()

# These are 'list' methods
meta.etl_jobs()
meta.data_sources()
meta.feature_sets(details=True/False)
meta.models(details=True/False)
meta.endpoints()
meta.views()
meta.pipelines()

# These are 'describe' methods
meta.data_source("abalone_data")
meta.feature_set("abalone_features")
meta.model("abalone-regression")
meta.endpoint("abalone-endpoint")
Source code in src/workbench/api/meta.py 
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
class Meta(CloudMeta):
    """Meta: A class that provides metadata functionality for Cloud Platform Artifacts.

    Common Usage:
       ```python
       from workbench.api import Meta
       meta = Meta()

       # Get the AWS Account Info
       meta.account()
       meta.config()

       # These are 'list' methods
       meta.etl_jobs()
       meta.data_sources()
       meta.feature_sets(details=True/False)
       meta.models(details=True/False)
       meta.endpoints()
       meta.views()
       meta.pipelines()

       # These are 'describe' methods
       meta.data_source("abalone_data")
       meta.feature_set("abalone_features")
       meta.model("abalone-regression")
       meta.endpoint("abalone-endpoint")
       ```
    """

    def account(self) -> dict:
        """Cloud Platform Account Info

        Returns:
            dict: Cloud Platform Account Info
        """
        return super().account()

    def config(self) -> dict:
        """Return the current Workbench Configuration

        Returns:
            dict: The current Workbench Configuration
        """
        return super().config()

    def incoming_data(self) -> pd.DataFrame:
        """Get summary data about data in the incoming raw data

        Returns:
            pd.DataFrame: A summary of the incoming raw data
        """
        return super().incoming_data()

    def etl_jobs(self) -> pd.DataFrame:
        """Get summary data about Extract, Transform, Load (ETL) Jobs

        Returns:
            pd.DataFrame: A summary of the ETL Jobs deployed in the Cloud Platform
        """
        return super().etl_jobs()

    def data_sources(self) -> pd.DataFrame:
        """Get a summary of the Data Sources deployed in the Cloud Platform

        Returns:
            pd.DataFrame: A summary of the Data Sources deployed in the Cloud Platform
        """
        return super().data_sources()

    def views(self, database: str = "workbench") -> pd.DataFrame:
        """Get a summary of the all the Views, for the given database, in AWS

        Args:
            database (str, optional): Glue database. Defaults to 'workbench'.

        Returns:
            pd.DataFrame: A summary of all the Views, for the given database, in AWS
        """
        return super().views(database=database)

    def feature_sets(self, details: bool = False) -> pd.DataFrame:
        """Get a summary of the Feature Sets deployed in the Cloud Platform

        Args:
            details (bool, optional): Include detailed information. Defaults to False.

        Returns:
            pd.DataFrame: A summary of the Feature Sets deployed in the Cloud Platform
        """
        return super().feature_sets(details=details)

    def models(self, details: bool = False) -> pd.DataFrame:
        """Get a summary of the Models deployed in the Cloud Platform

        Args:
            details (bool, optional): Include detailed information. Defaults to False.

        Returns:
            pd.DataFrame: A summary of the Models deployed in the Cloud Platform
        """
        return super().models(details=details)

    def endpoints(self) -> pd.DataFrame:
        """Get a summary of the Endpoints deployed in the Cloud Platform

        Returns:
            pd.DataFrame: A summary of the Endpoints in the Cloud Platform
        """
        return super().endpoints()

    def pipelines(self) -> pd.DataFrame:
        """Get a summary of the ML Pipelines deployed in the Cloud Platform

        Returns:
            pd.DataFrame: A summary of the Pipelines in the Cloud Platform
        """
        return super().pipelines()

    def glue_job(self, job_name: str) -> Union[dict, None]:
        """Get the details of a specific Glue Job

        Args:
            job_name (str): The name of the Glue Job

        Returns:
            dict: The details of the Glue Job (None if not found)
        """
        return super().glue_job(job_name=job_name)

    def data_source(self, data_source_name: str, database: str = "workbench") -> Union[dict, None]:
        """Get the details of a specific Data Source

        Args:
            data_source_name (str): The name of the Data Source
            database (str, optional): The Glue database. Defaults to 'workbench'.

        Returns:
            dict: The details of the Data Source (None if not found)
        """
        return super().data_source(table_name=data_source_name, database=database)

    def feature_set(self, feature_set_name: str) -> Union[dict, None]:
        """Get the details of a specific Feature Set

        Args:
            feature_set_name (str): The name of the Feature Set

        Returns:
            dict: The details of the Feature Set (None if not found)
        """
        return super().feature_set(feature_group_name=feature_set_name)

    def model(self, model_name: str) -> Union[dict, None]:
        """Get the details of a specific Model

        Args:
            model_name (str): The name of the Model

        Returns:
            dict: The details of the Model (None if not found)
        """
        return super().model(model_group_name=model_name)

    def endpoint(self, endpoint_name: str) -> Union[dict, None]:
        """Get the details of a specific Endpoint

        Args:
            endpoint_name (str): The name of the Endpoint

        Returns:
            dict: The details of the Endpoint (None if not found)
        """
        return super().endpoint(endpoint_name=endpoint_name)

    def __repr__(self):
        return f"Meta()\n\t{super().__repr__()}"

account()

Cloud Platform Account Info

Returns:

Name Type Description
dict dict

Cloud Platform Account Info
Source code in src/workbench/api/meta.py 
43
44
45
46
47
48
49
def account(self) -> dict:
    """Cloud Platform Account Info

    Returns:
        dict: Cloud Platform Account Info
    """
    return super().account()

config()

Return the current Workbench Configuration

Returns:

Name Type Description
dict dict

The current Workbench Configuration
Source code in src/workbench/api/meta.py 
51
52
53
54
55
56
57
def config(self) -> dict:
    """Return the current Workbench Configuration

    Returns:
        dict: The current Workbench Configuration
    """
    return super().config()

data_source(data_source_name, database='workbench')

Get the details of a specific Data Source

Parameters:

Name Type Description Default
data_source_name str

The name of the Data Source

 required
database str

The Glue database. Defaults to 'workbench'.

 'workbench'

Returns:

Name Type Description
dict Union[dict, None]

The details of the Data Source (None if not found)
Source code in src/workbench/api/meta.py 
143
144
145
146
147
148
149
150
151
152
153
def data_source(self, data_source_name: str, database: str = "workbench") -> Union[dict, None]:
    """Get the details of a specific Data Source

    Args:
        data_source_name (str): The name of the Data Source
        database (str, optional): The Glue database. Defaults to 'workbench'.

    Returns:
        dict: The details of the Data Source (None if not found)
    """
    return super().data_source(table_name=data_source_name, database=database)

data_sources()

Get a summary of the Data Sources deployed in the Cloud Platform

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the Data Sources deployed in the Cloud Platform
Source code in src/workbench/api/meta.py 
75
76
77
78
79
80
81
def data_sources(self) -> pd.DataFrame:
    """Get a summary of the Data Sources deployed in the Cloud Platform

    Returns:
        pd.DataFrame: A summary of the Data Sources deployed in the Cloud Platform
    """
    return super().data_sources()

endpoint(endpoint_name)

Get the details of a specific Endpoint

Parameters:

Name Type Description Default
endpoint_name str

The name of the Endpoint

 required

Returns:

Name Type Description
dict Union[dict, None]

The details of the Endpoint (None if not found)
Source code in src/workbench/api/meta.py 
177
178
179
180
181
182
183
184
185
186
def endpoint(self, endpoint_name: str) -> Union[dict, None]:
    """Get the details of a specific Endpoint

    Args:
        endpoint_name (str): The name of the Endpoint

    Returns:
        dict: The details of the Endpoint (None if not found)
    """
    return super().endpoint(endpoint_name=endpoint_name)

endpoints()

Get a summary of the Endpoints deployed in the Cloud Platform

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the Endpoints in the Cloud Platform
Source code in src/workbench/api/meta.py 
116
117
118
119
120
121
122
def endpoints(self) -> pd.DataFrame:
    """Get a summary of the Endpoints deployed in the Cloud Platform

    Returns:
        pd.DataFrame: A summary of the Endpoints in the Cloud Platform
    """
    return super().endpoints()

etl_jobs()

Get summary data about Extract, Transform, Load (ETL) Jobs

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the ETL Jobs deployed in the Cloud Platform
Source code in src/workbench/api/meta.py 
67
68
69
70
71
72
73
def etl_jobs(self) -> pd.DataFrame:
    """Get summary data about Extract, Transform, Load (ETL) Jobs

    Returns:
        pd.DataFrame: A summary of the ETL Jobs deployed in the Cloud Platform
    """
    return super().etl_jobs()

feature_set(feature_set_name)

Get the details of a specific Feature Set

Parameters:

Name Type Description Default
feature_set_name str

The name of the Feature Set

 required

Returns:

Name Type Description
dict Union[dict, None]

The details of the Feature Set (None if not found)
Source code in src/workbench/api/meta.py 
155
156
157
158
159
160
161
162
163
164
def feature_set(self, feature_set_name: str) -> Union[dict, None]:
    """Get the details of a specific Feature Set

    Args:
        feature_set_name (str): The name of the Feature Set

    Returns:
        dict: The details of the Feature Set (None if not found)
    """
    return super().feature_set(feature_group_name=feature_set_name)

feature_sets(details=False)

Get a summary of the Feature Sets deployed in the Cloud Platform

Parameters:

Name Type Description Default
details bool

Include detailed information. Defaults to False.

 False

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the Feature Sets deployed in the Cloud Platform
Source code in src/workbench/api/meta.py 
 94
 95
 96
 97
 98
 99
100
101
102
103
def feature_sets(self, details: bool = False) -> pd.DataFrame:
    """Get a summary of the Feature Sets deployed in the Cloud Platform

    Args:
        details (bool, optional): Include detailed information. Defaults to False.

    Returns:
        pd.DataFrame: A summary of the Feature Sets deployed in the Cloud Platform
    """
    return super().feature_sets(details=details)

glue_job(job_name)

Get the details of a specific Glue Job

Parameters:

Name Type Description Default
job_name str

The name of the Glue Job

 required

Returns:

Name Type Description
dict Union[dict, None]

The details of the Glue Job (None if not found)
Source code in src/workbench/api/meta.py 
132
133
134
135
136
137
138
139
140
141
def glue_job(self, job_name: str) -> Union[dict, None]:
    """Get the details of a specific Glue Job

    Args:
        job_name (str): The name of the Glue Job

    Returns:
        dict: The details of the Glue Job (None if not found)
    """
    return super().glue_job(job_name=job_name)

incoming_data()

Get summary data about data in the incoming raw data

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the incoming raw data
Source code in src/workbench/api/meta.py 
59
60
61
62
63
64
65
def incoming_data(self) -> pd.DataFrame:
    """Get summary data about data in the incoming raw data

    Returns:
        pd.DataFrame: A summary of the incoming raw data
    """
    return super().incoming_data()

model(model_name)

Get the details of a specific Model

Parameters:

Name Type Description Default
model_name str

The name of the Model

 required

Returns:

Name Type Description
dict Union[dict, None]

The details of the Model (None if not found)
Source code in src/workbench/api/meta.py 
166
167
168
169
170
171
172
173
174
175
def model(self, model_name: str) -> Union[dict, None]:
    """Get the details of a specific Model

    Args:
        model_name (str): The name of the Model

    Returns:
        dict: The details of the Model (None if not found)
    """
    return super().model(model_group_name=model_name)

models(details=False)

Get a summary of the Models deployed in the Cloud Platform

Parameters:

Name Type Description Default
details bool

Include detailed information. Defaults to False.

 False

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the Models deployed in the Cloud Platform
Source code in src/workbench/api/meta.py 
105
106
107
108
109
110
111
112
113
114
def models(self, details: bool = False) -> pd.DataFrame:
    """Get a summary of the Models deployed in the Cloud Platform

    Args:
        details (bool, optional): Include detailed information. Defaults to False.

    Returns:
        pd.DataFrame: A summary of the Models deployed in the Cloud Platform
    """
    return super().models(details=details)

pipelines()

Get a summary of the ML Pipelines deployed in the Cloud Platform

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of the Pipelines in the Cloud Platform
Source code in src/workbench/api/meta.py 
124
125
126
127
128
129
130
def pipelines(self) -> pd.DataFrame:
    """Get a summary of the ML Pipelines deployed in the Cloud Platform

    Returns:
        pd.DataFrame: A summary of the Pipelines in the Cloud Platform
    """
    return super().pipelines()

views(database='workbench')

Get a summary of the all the Views, for the given database, in AWS

Parameters:

Name Type Description Default
database str

Glue database. Defaults to 'workbench'.

 'workbench'

Returns:

Type Description
DataFrame

pd.DataFrame: A summary of all the Views, for the given database, in AWS
Source code in src/workbench/api/meta.py 
83
84
85
86
87
88
89
90
91
92
def views(self, database: str = "workbench") -> pd.DataFrame:
    """Get a summary of the all the Views, for the given database, in AWS

    Args:
        database (str, optional): Glue database. Defaults to 'workbench'.

    Returns:
        pd.DataFrame: A summary of all the Views, for the given database, in AWS
    """
    return super().views(database=database)

Examples

These example show how to use the Meta() class to pull lists of artifacts from AWS. DataSources, FeatureSets, Models, Endpoints and more. If you're building a web interface plugin, the Meta class is a great place to start.

Workbench REPL

If you'd like to see exactly what data/details you get back from the Meta() class, you can spin up the Workbench REPL, use the class and test out all the methods. Try it out! Workbench REPL

Using Workbench REPL
meta = Meta()
model_df = meta.models()
model_df
               Model Group   Health Owner  ...             Input     Status                Description
0      wine-classification  healthy     -  ...     wine_features  Completed  Wine Classification Model
1  abalone-regression-full  healthy     -  ...  abalone_features  Completed   Abalone Regression Model
2       abalone-regression  healthy     -  ...  abalone_features  Completed   Abalone Regression Model

[3 rows x 10 columns]

List the Models in AWS

meta_list_models.py
from workbench.api import Meta

# Create our Meta Class and get a list of our Models
meta = Meta()
model_df = meta.models()

print(f"Number of Models: {len(model_df)}")
print(model_df)

# Get more details data on the Models
model_names = model_df["Model Group"].tolist()
for name in model_names:
    pprint(meta.model(name))

Output

Number of Models: 3
               Model Group   Health Owner  ...             Input     Status                Description
0      wine-classification  healthy     -  ...     wine_features  Completed  Wine Classification Model
1  abalone-regression-full  healthy     -  ...  abalone_features  Completed   Abalone Regression Model
2       abalone-regression  healthy     -  ...  abalone_features  Completed   Abalone Regression Model

[3 rows x 10 columns]
wine-classification
abalone-regression-full
abalone-regression

Getting Model Performance Metrics

meta_model_metrics.py
from workbench.api import Meta

# Create our Meta Class and get a list of our Models
meta = Meta()
model_df = meta.models()

print(f"Number of Models: {len(model_df)}")
print(model_df)

# Get more details data on the Models
model_names = model_df["Model Group"].tolist()
for name in model_names[:5]:
    model_details = meta.model(name)
    print(f"\n\nModel: {name}")
    performance_metrics = model_details["workbench_meta"]["workbench_inference_metrics"]
    print(f"\tPerformance Metrics: {performance_metrics}")

Output

wine-classification
    ARN: arn:aws:sagemaker:us-west-2:507740646243:model-package-group/wine-classification
    Description: Wine Classification Model
    Tags: wine::classification
    Performance Metrics:
        [{'wine_class': 'TypeA', 'precision': 1.0, 'recall': 1.0, 'fscore': 1.0, 'roc_auc': 1.0, 'support': 12}, {'wine_class': 'TypeB', 'precision': 1.0, 'recall': 1.0, 'fscore': 1.0, 'roc_auc': 1.0, 'support': 14}, {'wine_class': 'TypeC', 'precision': 1.0, 'recall': 1.0, 'fscore': 1.0, 'roc_auc': 1.0, 'support': 9}]

abalone-regression
    ARN: arn:aws:sagemaker:us-west-2:507740646243:model-package-group/abalone-regression
    Description: Abalone Regression Model
    Tags: abalone::regression
    Performance Metrics:
        [{'MAE': 1.64, 'RMSE': 2.246, 'R2': 0.502, 'MAPE': 16.393, 'MedAE': 1.209, 'NumRows': 834}]

List the Endpoints in AWS

meta_list_endpoints.py
from pprint import pprint
from workbench.api import Meta

# Create our Meta Class and get a list of our Endpoints
meta = Meta()
endpoint_df = meta.endpoints()
print(f"Number of Endpoints: {len(endpoint_df)}")
print(endpoint_df)

# Get more details data on the Endpoints
endpoint_names = endpoint_df["Name"].tolist()
for name in endpoint_names:
    pprint(meta.endpoint(name))

Output

Number of Endpoints: 2
                      Name   Health            Instance           Created  ...     Status     Variant Capture Samp(%)
0  wine-classification-end  healthy  Serverless (2GB/5)  2024-03-23 23:09  ...  InService  AllTraffic   False       -
1   abalone-regression-end  healthy  Serverless (2GB/5)  2024-03-23 21:11  ...  InService  AllTraffic   False       -

[2 rows x 10 columns]
wine-classification-end
<lots of details about endpoints>

Not Finding some particular AWS Data?

The Workbench Meta API Class also has (details=True) arguments, so make sure to check those out.