Example usage¶

To use aiod in a project:

In [1]:

import aiod

import aiod

In [2]:

aiod.config.api_base_url = "http://localhost/"
aiod.config.auth_server_url = "http://localhost/aiod-auth/"

aiod.config.api_base_url = "http://localhost/" aiod.config.auth_server_url = "http://localhost/aiod-auth/"

In [3]:

aiod_counts = aiod.counts()
aiod_counts[["datasets", "ml_models"]]

aiod_counts = aiod.counts() aiod_counts[["datasets", "ml_models"]]

Out[3]:

	datasets	ml_models
bonseyes	NaN	12.0
example	NaN	1.0
huggingface	122252.0	1.0
openml	5569.0	15767.0
stairwai	1.0	NaN
zenodo	305140.0	NaN
aida_cms	NaN	NaN
aiod	NaN	NaN
robotics4eu	NaN	NaN

In [3]:

help(aiod.datasets)

help(aiod.datasets)

Help on module aiod_sdk.resources.datasets in aiod_sdk.resources:

NAME
    aiod_sdk.resources.datasets

DATA
    counts = functools.partial(<function counts at 0x1062d54e0>, asset_typ...
        Retrieve the number of datasets assets in the metadata catalogue.
        
        Parameters (keywords required):
            version (str | None): The version of the endpoint (default is None).
            per_platform (bool): Whether to list counts per platform (default is False).
        
        Returns:
            int | dict[str, int]: The number datasets assets in the metadata catalogue.
                If the parameter per_platform is True, it returns a dict[str, int].
    
    get_asset = functools.partial(<function get_asset at 0x1185527a0>, ass...
        Retrieve metadata for a specific datasets.
        
        Parameters (keywords required):
            identifier (int): The identifier of the datasets to retrieve.
            version (str | None): The version of the endpoint (default is None).
            data_format (Literal["pandas", "json"]): The desired format for the response (default is "pandas").
                For "json" formats, the returned type is a json decoded type, in this case a dict.
        
        Returns:
            pd.Series | dict: The retrieved metadata for the specified datasets.
    
    get_assets_async = functools.partial(<function get_assets_async at 0x1...
        Asynchronously retrieve metadata for a list of datasets identifiers.
        
        Parameters (keywords required):
            identifiers (list[int]): The list of identifiers of the datasets to retrieve.
            version (str | None): The version of the endpoint (default is None).
            data_format (Literal["pandas", "json"]): The desired format for the response (default is "pandas").
                For "json" formats, the returned type is a json decoded type, in this case a list of dict's.
        
        Returns:
            pd.DataFrame | list[dict]: The retrieved metadata for the specified datasets.
    
    get_list = functools.partial(<function get_list at 0x1060365c0>, asset...
        Retrieve a list of datasets from the catalogue.
        
        Parameters (keywords required):
            offset (int): The offset for pagination (default is 0).
            limit (int): The maximum number of items to retrieve (default is 10).
            version (str | None): The version of the endpoint (default is None).
            data_format (Literal["pandas", "json"]): The desired format for the response (default is "pandas").
                For "json" formats, the returned type is a json decoded type, i.e. in this case a list of dict's.
        
        Returns:
            pd.DataFrame | list[dict]: The retrieved metadata in the specified format.
    
    get_list_async = functools.partial(<function get_list_async at 0x11855...
        Asynchronously retrieve a list of datasets from the catalogue in batches.
        
        Parameters (keywords required):
            offset (int): The offset for pagination (default is 0).
            limit (int): The maximum number of items to retrieve (default is 10).
            batch_size (int): The number of items in a a batch.
            version (str | None): The version of the endpoint (default is None).
            data_format (Literal["pandas", "json"]): The desired format for the response (default is "pandas").
                For "json" formats, the returned type is a json decoded type, in this case a list of dict's.
        
        Returns:
            pd.DataFrame | list[dict]: The retrieved metadata in the specified format.

FILE
    /Users/jmatias/Documents/develop/aiod-py-sdk/src/aiod_sdk/resources/datasets.py

In [4]:

aiod.datasets.counts()

aiod.datasets.counts()

Out[4]:

41114

In [5]:

datasets = aiod.datasets.get_list(offset=300, limit=2)
datasets[["platform", "platform_resource_identifier", "name", "date_published"]]

datasets = aiod.datasets.get_list(offset=300, limit=2) datasets[["platform", "platform_resource_identifier", "name", "date_published"]]

Out[5]:

	platform	platform_resource_identifier	name	date_published
0	openml	436	rosowky	2014-09-28T23:37:11
1	openml	437	garrat	2014-09-28T23:37:13

In [6]:

metadata_dataset302 = aiod.datasets.get_asset(identifier=302)
metadata_dataset302

metadata_dataset302 = aiod.datasets.get_asset(identifier=302) metadata_dataset302

Out[6]:

platform                                                                   openml
platform_resource_identifier                                                  437
name                                                                       garrat
date_published                                                2014-09-28T23:37:13
same_as                               https://www.openml.org/api/v1/json/data/437
is_accessible_for_free                                                       True
version                                                                         1
ai_asset_identifier                                                           307
ai_resource_identifier                                                        317
aiod_entry                      {'editor': [], 'status': 'published', 'date_mo...
alternate_name                                                                 []
application_area                                                               []
citation                                                                       []
contact                                                                        []
creator                                                                        []
description                     {'plain': '**Author**:   
**Source**: Unknown ...
distribution                    [{'content_url': 'https://api.openml.org/data/...
funder                                                                         []
has_part                                                                       []
industrial_sector                                                              []
is_part_of                                                                     []
keyword                                           [machine learning, mathematics]
license                                                                    public
media                                                                          []
note                                                                           []
relevant_link                                                                  []
relevant_resource                                                              []
relevant_to                                                                    []
research_area                                                                  []
scientific_domain                                                              []
size                                           {'unit': 'instances', 'value': 10}
identifier                                                                    302
dtype: object

Async methods¶

In [7]:

data = await aiod.datasets.get_list_async(offset=300, limit=5, batch_size=3)
data[["platform", "name", "identifier", "is_accessible_for_free"]]

data = await aiod.datasets.get_list_async(offset=300, limit=5, batch_size=3) data[["platform", "name", "identifier", "is_accessible_for_free"]]

Out[7]:

	platform	name	identifier	is_accessible_for_free
0	openml	rosowky	301	True
1	openml	garrat	302	True
2	openml	doherty	303	True
3	openml	chang	304	True
4	openml	qsabr2	305	True

Authentication¶

In [11]:

aiod.config.api_base_url = "https://aiod-dev.i3a.es/"
aiod.config.auth_server_url = "https://aiod-dev.i3a.es/aiod-auth/"

aiod.config.api_base_url = "https://aiod-dev.i3a.es/" aiod.config.auth_server_url = "https://aiod-dev.i3a.es/aiod-auth/"

In [9]:

aiod.login(username="USERNAME", password="PASSWORD")
# aiod.get_current_user()

aiod.login(username="USERNAME", password="PASSWORD") # aiod.get_current_user()

In [9]:

aiod.logout()

aiod.logout()

In [12]:

aiod.get_current_user()

aiod.get_current_user()

Out[12]:

User(name='jsmatias', roles=('offline_access', 'uma_authorization', 'default-roles-aiod'))

In [ ]: