To help you get started, we've collected the most common ways that bigml is being used within popular public projects.

BigML Python Bindings

BigML makes machine learning easy by taking care of the details required to add data-driven decisions and predictive power to your company. Unlike other machine learning services, BigML creates beautiful predictive models that can be easily understood and interacted with.

These BigML Python bindings allow you to interact with BigML.io, the API for BigML. You can use it to easily create, retrieve, list, update, and delete BigML resources (i.e., sources, datasets, models and, predictions). For additional information, see the full documentation for the Python bindings on Read the Docs.

This module is licensed under the Apache License, Version 2.0.

Support

Please report problems and bugs to our BigML.io issue tracker.

Discussions about the different bindings take place in the general BigML mailing list. Or join us in our Campfire chatroom.

Requirements

Only Python 3 versions are currently supported by these bindings. Support for Python 2.7.X ended in version 4.32.3.

The basic third-party dependencies are the requests, unidecode, requests-toolbelt, bigml-chronos, msgpack, numpy and scipy libraries. These libraries are automatically installed during the basic setup. Support for Google App Engine has been added as of version 3.0.0, using the urlfetch package instead of requests.

The bindings will also use simplejson if you happen to have it installed, but that is optional: we fall back to Python's built-in JSON libraries is simplejson is not found.

The bindings provide support to use the BigML platform to create, update, get and delete resources, but also to produce local predictions using the models created in BigML. Most of them will be actionable with the basic installation, but some additional dependencies are needed to use local Topic Models and Image Processing models. Please, refer to the Installation section for details.

OS Requirements

The basic installation of the bindings is compatible and can be used on Linux and Windows based Operating Systems. However, the extra options that allow working with image processing models ([images] and [full]) are only supported and tested on Linux-based Operating Systems. For image models, Windows OS is not recommended and cannot be supported out of the box, because the specific compiler versions or dlls required are unavailable in general.

Installation

To install the basic latest stable release with pip, please use:

$ pip install bigml

Support for local Topic Distributions (Topic Models' predictions) and local predictions for datasets that include Images will only be available as extras, because the libraries used for that are not usually available in all Operative Systems. If you need to support those, please check the Installation Extras section.

Installation Extras

Local Topic Distributions support can be installed using:

pip install bigml[topics]

Images local predictions support can be installed using:

pip install bigml[images]

The full set of features can be installed using:

pip install bigml[full]

WARNING: Mind that installing these extras can require some extra work, as explained in the Requirements section.

You can also install the development version of the bindings directly from the Git repository

$ pip install -e git://github.com/bigmlcom/python.git#egg=bigml_python

Running the Tests

The tests will be run using pytest. You'll need to set up your authentication via environment variables, as explained in the authentication section. Also some of the tests need other environment variables like BIGML_ORGANIZATION to test calls when used by Organization members and BIGML_EXTERNAL_CONN_HOST, BIGML_EXTERNAL_CONN_PORT, BIGML_EXTERNAL_CONN_DB, BIGML_EXTERNAL_CONN_USER, BIGML_EXTERNAL_CONN_PWD and BIGML_EXTERNAL_CONN_SOURCE in order to test external data connectors.

With that in place, you can run the test suite simply by issuing

$ pytest

Additionally, Tox can be used to automatically run the test suite in virtual environments for all supported Python versions. To install Tox:

$ pip install tox

Then run the tests from the top-level project directory:

$ tox

Importing the module

To import the module:

import bigml.api

Alternatively you can just import the BigML class:

from bigml.api import BigML

Authentication

All the requests to BigML.io must be authenticated using your username and API key and are always transmitted over HTTPS.

This module will look for your username and API key in the environment variables BIGML_USERNAME and BIGML_API_KEY respectively.

Unix and MacOS

You can add the following lines to your .bashrc or .bash_profile to set those variables automatically when you log in:

export BIGML_USERNAME=myusername
export BIGML_API_KEY=ae579e7e53fb9abd646a6ff8aa99d4afe83ac291

refer to the next chapters to know how to do that in other operating systems.

With that environment set up, connecting to BigML is a breeze:

from bigml.api import BigML
api = BigML()

Otherwise, you can initialize directly when instantiating the BigML class as follows:

api = BigML('myusername', 'ae579e7e53fb9abd646a6ff8aa99d4afe83ac291')

These credentials will allow you to manage any resource in your user environment.

In BigML a user can also work for an organization. In this case, the organization administrator should previously assign permissions for the user to access one or several particular projects in the organization. Once permissions are granted, the user can work with resources in a project according to his permission level by creating a special constructor for each project. The connection constructor in this case should include the project ID:

api = BigML('myusername', 'ae579e7e53fb9abd646a6ff8aa99d4afe83ac291',
            project='project/53739b98d994972da7001d4a')

If the project used in a connection object does not belong to an existing organization but is one of the projects under the user's account, all the resources created or updated with that connection will also be assigned to the specified project.

When the resource to be managed is a project itself, the connection needs to include the correspondingorganization ID:

api = BigML('myusername', 'ae579e7e53fb9abd646a6ff8aa99d4afe83ac291',
            organization='organization/53739b98d994972da7025d4a')

Authentication on Windows

The credentials should be permanently stored in your system using

setx BIGML_USERNAME myusername
setx BIGML_API_KEY ae579e7e53fb9abd646a6ff8aa99d4afe83ac291

Note that setx will not change the environment variables of your actual console, so you will need to open a new one to start using them.

Authentication on Jupyter Notebook

You can set the environment variables using the %env command in your cells:

%env BIGML_USERNAME=myusername
%env BIGML_API_KEY=ae579e7e53fb9abd646a6ff8aa99d4afe83ac291

Alternative domains

The main public domain for the API service is bigml.io, but there are some alternative domains, either for Virtual Private Cloud setups or the australian subdomain (au.bigml.io). You can change the remote server domain to the VPC particular one by either setting the BIGML_DOMAIN environment variable to your VPC subdomain:

export BIGML_DOMAIN=my_VPC.bigml.io

or setting it when instantiating your connection:

api = BigML(domain="my_VPC.bigml.io")

The corresponding SSL REST calls will be directed to your private domain henceforth.

You can also set up your connection to use a particular PredictServer only for predictions. In order to do so, you'll need to specify a Domain object, where you can set up the general domain name as well as the particular prediction domain name.

from bigml.domain import Domain
from bigml.api import BigML

domain_info = Domain(prediction_domain="my_prediction_server.bigml.com",
                     prediction_protocol="http")

api = BigML(domain=domain_info)

Finally, you can combine all the options and change both the general domain server, and the prediction domain server.

from bigml.domain import Domain
from bigml.api import BigML
domain_info = Domain(domain="my_VPC.bigml.io",
                     prediction_domain="my_prediction_server.bigml.com",
                     prediction_protocol="https")

api = BigML(domain=domain_info)

Some arguments for the Domain constructor are more unsual, but they can also be used to set your special service endpoints:

• protocol (string) Protocol for the service (when different from HTTPS)
• verify (boolean) Sets on/off the SSL verification
• prediction_verify (boolean) Sets on/off the SSL verification for the prediction server (when different from the general SSL verification)

Note that the previously existing dev_mode flag:

api = BigML(dev_mode=True)

that caused the connection to work with the Sandbox Development Environment has been deprecated because this environment does not longer exist. The existing resources that were previously created in this environment have been moved to a special project in the now unique Production Environment, so this flag is no longer needed to work with them.

Quick Start

Imagine that you want to use this csv file containing the Iris flower dataset to predict the species of a flower whose petal length is 2.45 and whose petal width is 1.75. A preview of the dataset is shown below. It has 4 numeric fields: sepal length, sepal width, petal length, petal width and a categorical field: species. By default, BigML considers the last field in the dataset as the objective field (i.e., the field that you want to generate predictions for).

sepal length,sepal width,petal length,petal width,species
5.1,3.5,1.4,0.2,Iris-setosa
4.9,3.0,1.4,0.2,Iris-setosa
4.7,3.2,1.3,0.2,Iris-setosa
...
5.8,2.7,3.9,1.2,Iris-versicolor
6.0,2.7,5.1,1.6,Iris-versicolor
5.4,3.0,4.5,1.5,Iris-versicolor
...
6.8,3.0,5.5,2.1,Iris-virginica
5.7,2.5,5.0,2.0,Iris-virginica
5.8,2.8,5.1,2.4,Iris-virginica

You can easily generate a prediction following these steps:

from bigml.api import BigML

api = BigML()

source = api.create_source('./data/iris.csv')
dataset = api.create_dataset(source)
model = api.create_model(dataset)
prediction = api.create_prediction(model, \
    {"petal width": 1.75, "petal length": 2.45})

You can then print the prediction using the pprint method:

>>> api.pprint(prediction)
species for {"petal width": 1.75, "petal length": 2.45} is Iris-setosa

Certainly, any of the resources created in BigML can be configured using several arguments described in the API documentation. Any of these configuration arguments can be added to the create method as a dictionary in the last optional argument of the calls:

from bigml.api import BigML

api = BigML()

source_args = {"name": "my source",
     "source_parser": {"missing_tokens": ["NULL"]}}
source = api.create_source('./data/iris.csv', source_args)
dataset_args = {"name": "my dataset"}
dataset = api.create_dataset(source, dataset_args)
model_args = {"objective_field": "species"}
model = api.create_model(dataset, model_args)
prediction_args = {"name": "my prediction"}
prediction = api.create_prediction(model, \
    {"petal width": 1.75, "petal length": 2.45},
    prediction_args)

The iris dataset has a small number of instances, and usually will be instantly created, so the api.create_ calls will probably return the finished resources outright. As BigML's API is asynchronous, in general you will need to ensure that objects are finished before using them by using api.ok.

from bigml.api import BigML

api = BigML()

source = api.create_source('./data/iris.csv')
api.ok(source)
dataset = api.create_dataset(source)
api.ok(dataset)
model = api.create_model(dataset)
api.ok(model)
prediction = api.create_prediction(model, \
    {"petal width": 1.75, "petal length": 2.45})

Note that the prediction call is not followed by the api.ok method. Predictions are so quick to be generated that, unlike the rest of resouces, will be generated synchronously as a finished object.

The example assumes that your objective field (the one you want to predict) is the last field in the dataset. If that's not he case, you can explicitly set the name of this field in the creation call using the objective_field argument:

from bigml.api import BigML

api = BigML()

source = api.create_source('./data/iris.csv')
api.ok(source)
dataset = api.create_dataset(source)
api.ok(dataset)
model = api.create_model(dataset, {"objective_field": "species"})
api.ok(model)
prediction = api.create_prediction(model, \
    {'sepal length': 5, 'sepal width': 2.5})

You can also generate an evaluation for the model by using:

test_source = api.create_source('./data/test_iris.csv')
api.ok(test_source)
test_dataset = api.create_dataset(test_source)
api.ok(test_dataset)
evaluation = api.create_evaluation(model, test_dataset)
api.ok(evaluation)

If you set the storage argument in the api instantiation:

api = BigML(storage='./storage')

all the generated, updated or retrieved resources will be automatically saved to the chosen directory.

Alternatively, you can use the export method to explicitly download the JSON information that describes any of your resources in BigML to a particular file:

api.export('model/5acea49a08b07e14b9001068',
           filename="my_dir/my_model.json")

This example downloads the JSON for the model and stores it in the my_dir/my_model.json file.

In the case of models that can be represented in a PMML syntax, the export method can be used to produce the corresponding PMML file.

api.export('model/5acea49a08b07e14b9001068',
           filename="my_dir/my_model.pmml",
           pmml=True)

You can also retrieve the last resource with some previously given tag:

api.export_last("foo",
                resource_type="ensemble",
                filename="my_dir/my_ensemble.json")

which selects the last ensemble that has a foo tag. This mechanism can be specially useful when retrieving retrained models that have been created with a shared unique keyword as tag.

For a descriptive overview of the steps that you will usually need to follow to model your data and obtain predictions, please see the basic Workflow sketch document. You can also check other simple examples in the following documents:

• model 101
• logistic regression 101
• linear regression 101
• ensemble 101
• cluster 101
• anomaly detector 101
• association 101
• topic model 101
• deepnet 101
• time series 101
• fusion 101
• scripting 101

Additional Information

We've just barely scratched the surface. For additional information, see the full documentation for the Python bindings on Read the Docs. Alternatively, the same documentation can be built from a local checkout of the source by installing Sphinx ($ pip install sphinx) and then running

$ cd docs
$ make html

Then launch docs/_build/html/index.html in your browser.

How to Contribute

Please follow the next steps:

> 1. Fork the project on github.com. > 2. Create a new branch. > 3. Commit changes to the new branch. > 4. Send a pull request.

For details on the underlying API, see the BigML API documentation.

History

9.6.2 (2023-08-02)

• Extending cloning to all available models and WhizzML scripts.
• Fixing shared resources cloning.

9.6.1 (2023-08-01)

• Adding shared resources cloning.

9.6.0 (2023-07-20)

• Adding ShapWrapper to enable local Shap values computation with the Shap library.
• Adding Evaluation object.
• Improving Field class to allow field values encoding as numpy arrays.

9.5.0 (2023-06-16)

• Extending Local Fusions output to include confidence.

9.4.0 (2023-06-14)

• Extending LocalModel class to handle Time Series locally.

9.3.0 (2023-06-09)

• Adding a LocalModel class to handle any type of BigML model locally.

9.2.0 (2023-04-11)

• Extending all delete methods to allow additional query strings.

9.1.4 (2023-02-10)

• Providing local deepnet predictions default for Windows OS that cannot handle images predictions.

9.1.3 (2022-12-22)

• Changing user's status endpoint retrieval to a lightweight version.

9.1.2 (2022-11-26)

• Removing all nose dependencies in tests.

9.1.1 (2022-11-18)

• Removing traces and refactoring Flatline interpreter invocation.
• Migrating tests to pytest.

9.1.0 (2022-11-09)

• Enabling local models to generate the transformations pipeline used in training to be able to apply it locally to input data for predictions.
• Refactoring local models to bring the feature extraction capabilities to the transformations pipeline.
• Rounding local numeric predictions and scores to 5 digits.
• Improving Pipelines and Image Processing documentation.

9.0.1 (2022-11-02)

• Fixing local batch_prediction method on DataFrames.
• Improving the training examples method in the Fields class to produce examples of categorical field values following their original distribution.

9.0.0 (2022-10-12)

• Renaming Pipeline class to BMLPipeline and refactoring to allow scikit-learn and Pandas pipelines to be also part of the Pipeline transformations.
• Adding several DataTransformers (BMLDataTransformer, SKDataTransformer and DFDataTransformer) as wrappers for transformation generators.

8.2.3 (2022-10-07)

• Fixing dump methods for all local resources.
• Refactoring output options in batch_predict methods.
• Adding name and description to local resources.

8.2.2 (2022-09-29)

• Fixing locale check.
• Documenting the new .batch_predict method added to local models to homogenize local batch predictions and accept Pandas' DataFrame as input.

8.2.1 (2022-09-23)

• Fixing update annotations function when creating annotated images sources.

8.2.0 (2022-09-22)

• Adding new option to api.ok to report progress via a function callback.

8.1.1 (2022-09-21)

• Fixing bug in api.ok: response to failing API calls.

8.1.0 (2022-09-16)

• Upgrading dependencies.

8.0.0 (2022-09-13)

• Adding new local Pipeline object to encapsulate and run prediction workflows.
• Changing api.ok response to return as False when retrieving faulty resources. Previously, only problems retrieving the API response cause api.ok returning False.
• Adding FaultyResource exception to inform about that type of failure when using the raise_on_error flag in api.ok.
• Fixing bug in local LogisticRegression when predicting with operating points.

7.7.0 (2022-08-30)

• Adding local Dataset class to handle Flatline transformations.

7.6.0 (2022-08-18)

• Adding feature expansion to image fields in local predictors.

7.5.1 (2022-08-08)

• Adding utility to reproduce the backend images preprocess in local deepnets.

7.5.0 (2022-07-09)

• Adding the new capabilities the last API release: object detection.

7.4.2 (2022-05-26)

• Improved webhooks hashing check.
• Adapting connection to allow API version setting, needed for compatibility with the Predict Server syntax.

7.4.1 (2022-05-16)

• Adding support for webhooks hashing check.

7.4.0 (2022-05-13)

• Adding extras to setup, to allow a basic installation and extra dependencies handling for topics and images.

7.3.0 (2022-04-22)

• Improving local ensembles to be generated from list of ensemble plus models info.
• Allowing BigML connection class to retrieve models from storage without any credentials.

7.2.2 (2022-04-20)

• Enabling dump from SupervisedModel class.

7.2.1 (2022-04-19)

• Fixing windows setup.

7.2.0 (2022-04-14)

• Upgrading sensenet library and giving fallback for local deepnet predictions on OS not supported by tensorflow.
• Adding support for pandas' DataFrame when creating sources.

7.1.2 (2022-03-17)

• Fixing local deepnet predictions if no missing_count info is found.

7.1.1 (2022-03-11)

• Improving local predictions for shared fusions and documenting.

7.1.0 (2022-03-10)

• Adding the ability to predict locally from shared composed resources, like Ensembles and Fusions.

7.0.1 (2022-03-04)

• Fixing local Supervised Model to accept stored models as input.
• Improving tests to minimize resource creation.
• Updating dependencies versions.

7.0.0 (2021-08-03)

• Adding composite sources to the API REST interface methods.
• Adding predictions for images.

6.3.2 (2021-12-02)

• Fixing local deepnet prediction method to fit new deepnet structure.

6.3.1 (2021-11-03)

• Improving local ensembles to handle faulty models with no root.

6.3.0 (2021-10-21)

• Adding activators to local deepnets prediction functions.

6.2.2 (2021-10-11)

• Fixing underlying dependency used in datetime fields.

6.2.1 (2021-08-10)

• Fixing predictions for local models with text fields.

6.2.0 (2021-07-22)

• Minimizing local model information when text or items fields are used.

6.1.2 (2021-06-16)

• Changing tests mechanism to github actions.

6.1.1 (2021-06-11)

• Fixing api get method. The query string argument was lost.
• Improving api get, create and update methods to avoid unneeded iterations.
• Improving docs.

6.1.0 (2021-03-06)

• Fixing local anomalies when normalize_repeats has been used.

6.0.3 (2021-02-06)

• Improving local models with default numeric values to automatically fill in missings in input data.
• Raising exceptions if trying to update a failed resource.

6.0.2 (2020-12-21)

• Fixing bug in local anomaly filter.

6.0.1 (2020-11-11)

• Changing local anomaly filter.
• Fixing update and download methods to wait for the resource to be finished.

6.0.0 (2020-09-02)

• New compact format for local models.
• Adding cache managing capacities for all local predictors.

5.2.0 (2020-08-19)

• Adding syntactic sugar methods to the BigML class to help managing resources.

5.1.1 (2020-08-11)

• Fixing module directory inclusion and improving docs on local anomalies.

5.1.0 (2020-08-07)

• Refactoring local anomaly to reduce memory requirements.

5.0.1 (2020-08-05)

• Fixing bug in get_tasks_status to get information about transient net errors.

5.0.0 (2020-07-31)

• Deprecating support for Python 2.7.X versions. Only Python 3 supported from this version on.

4.32.3 (2020-07-15)

• Extending the Fields class to check the attributes that can be updated in a source or dataset fields structure to avoid failing fields updates.

4.32.2 (2020-06-15)

• Fixing local anomaly scores for new anomaly detectors with feedback and setting the maximum input data precision to five digits.

4.32.1 (2020-06-10)

• Fixing local anomaly scores prediction for corner cases of samples with one row.

4.32.0 (2020-05-19)

• Allowing scripts to be created from gists using the create_script method.
• Improving training examples generation in Fields class.

4.31.2 (2020-05-14)

• Fixing problems creating ephemeral prediction resources.

4.31.1 (2020-05-06)

• Improving the api.ok method to add an estimated wait time.
• Improving docs and adding TOC for new structure.

4.31.0 (2020-04-22)

• Adding REST methods to manage external data connections.

4.30.2 (2020-04-20)

• Fixing local anomaly scores for datasets with significant amounts of missings.
• Fixing input data modification for local predictions when fields are not used in the models.

4.30.1 (2020-04-16)

• Fixing tasks status info for organizations.

4.30.0 (2020-04-10)

• Allowing the BigML class to retrieve any resource from local storage and extract its fields.

4.29.2 (2020-03-20)

• Improving exception handling when retrieving resources.

4.29.1 (2020-03-03)

• Fixing bug when disabling SSL verification in predictions only.

4.29.0 (2020-02-29)

• Improving api.ok method to allow retries to avoid transient HTTP failures.
• Deprecating the retries argument in api.ok.
• Fixing local predictions confidence for weighted models.

4.28.1 (2020-02-04)

• Changing api.ok method to avoid raising exceptions when retrieving a faulty resource.
• Adding call stack info to local Execution class.
• Fixing docs builder.

4.28.0 (2020-01-23)

• Adding Execution local utility to extract the outputs and results from an execution.

4.27.3 (2020-01-15)

• Fixing local Fusion class to allow using linear regressions.

4.27.2 (2020-01-03)

• Fixing warning message and template files in generated code for hadoop actionable models.
• Fixing local ensembles that asked for credentials before needing them.

4.27.1 (2019-12-19)

• Avoiding asking for credential in classes that predict locally when the complete information is provided so no connection is needed.

4.27.0 (2019-12-03)

• Extending the custom formats for datetimes allowed as input for local predictions.
• Fixing datetimes allowed as input for local predictions. They can be provided by name or ID.

4.26.0 (2019-11-27)

• Extending the ability to use an alternative url to all predictions, centroids, anomaly scores, etc. Also to their batch versions.

4.25.3 (2019-11-26)

• Changing bigml-chronos dependency version according to its new internal structure. The previous version caused problems when used in some external projects.

4.25.2 (2019-11-06)

• Fixing bug in local Cluster object when using text or item fields.

4.25.1 (2019-08-28)

• Fixing bug in local Fusion object when retrienving from storage.

4.25.0 (2019-08-18)

• Adding the ability to parse datetime fields locally for local predictions (uses bigml-chronos as a dependency).

4.24.3 (2019-08-08)

• Fixing local LinearRegression to work even if numpy and scipy are not installed.

4.24.2 (2019-07-30)

• Fixing local EnsemblePredictor code to avoid crash when using deep trees.

4.24.1 (2019-07-05)

• Adding missing tokens handling to local models.

4.24.0 (2019-06-28)

• Refactoring for multipackage compatibility.
• Deprecating ensemble_id attribute in local ensembles.
• Extending the BigML class to export model's alternative output formats.

4.23.1 (2019-06-06)

• Fixing local predictions for models with unpreferred and datetime fields.

4.23.0 (2019-05-24)

• Adding access to tasks information in the API connection object.

4.22.1 (2019-05-23)

• Improving the local Ensemble and Fusion classes to use the component models when a local JSON file is used as argument.

4.22.0 (2019-05-11)

• Fixing bug in local linear regressions for non-invertible confidence bounds matrices.
• Adding the option of cloning model resources from shared clonable ones.
• Fixing Fields object for timeseries.

4.21.2 (2019-04-09)

• Fixing bug in local fusion regression predictions.

4.21.1 (2019-04-06)

• Fixing bug in local linear regression predictions.

4.21.0 (2019-03-22)

• Adding REST and local methods for linear regression.

4.20.2 (2019-02-02)

• Adding new format for the list of datasets to create a multidataset from.

4.20.1 (2019-02-01)

• Fixing bug in local ensemble when used with externally defined connection, as found by @KamalGalrani.

4.20.0 (2018-12-01)

• Adding PCA REST call methods.
• Adding local PCAs and Projections.

4.19.10 (2018-12-01)

• Fixing local Deepnet predictions for regressions without numpy.

4.19.9 (2018-10-24)

• Fixing bug in create datasets for a list of one dataset only.

4.19.8 (2018-09-18)

• Fixing bug in create evaluation for timeseries.

4.19.7 (2018-09-13)

• Fixing bug when exporting fusions with weights.
• Local fusions now caching all models in the constructor.

4.19.6 (2018-09-12)

• Fixing bug when exporting fusions.

4.19.5 (2018-08-23)

• Changing source upload async parameter to ensure Python 3.7 compatibility.

4.19.4 (2018-07-18)

• Fixing local logistic regression predictions with weight field missing in input data.

4.19.3 (2018-06-26)

• Modifying local fusion object to adapt to logistic regressions with no missing numerics allowed.

4.19.2 (2018-06-25)

• Removing left over comment.

4.19.1 (2018-06-23)

• Refactoring the local classes that manage models information to create predictions. Now all of them allow a path, an ID or a dictionary to be the first argument in the constructor.

4.19.0 (2018-06-20)

• Adding local fusion object and predict methods.
• Fixing error handling in local objects.
• Fixing bug in local logistic regressions when using a local stored file.

4.18.3 (2018-06-03)

• Adding batch predictions for fusion resources.

4.18.2 (2018-05-28)

• Adding predictions and evaluations for fusion resources.

4.18.1 (2018-05-19)

• Fixing bug when unused field IDs are used in local prediction inputs.

4.18.0 (2018-05-19)

• Adding methods for the REST calls to OptiMLs and Fusions.

4.17.1 (2018-05-15)

• Adding the option to export PMML models when available.
• Fixing bug in local deepnets for regressions.
• Adapting local Cluster and Anomaly detector to not include summary fields information.

4.17.0 (2018-05-02)

• Adding the local Supervised Model class to allow local predictions with any supervised model resource.

4.16.2 (2018-04-31)

• Adding the export and export_last methods to download and save the remote resources in the local file system.

4.16.1 (2018-04-24)

• Fixing bug in local deepnet predictions.

4.16.0 (2018-04-03)

• Deprecating local predictions formatting arguments. Formatting is available through the cast_prediction function.

4.15.2 (2018-02-24)

• Local predictions for regression ensembles corrected for strange models whose nodes lack the confidence attribute.

4.15.1 (2018-02-07)

• Removing logs left in local ensemble object.

4.15.0 (2018-02-07)

• Adding organizations support for all the API calls.

4.14.0 (2018-01-22)

• Deprecating dev_mode flag from BigML's API connection. The development environment has been deprecated.
• Fixing bug in local cluster output to CSV.
• Improving docs with local batch predictions examples.
• Adding operating kind support for local predictions in models and ensembles.
• Fixing bug in ensembles local predictions with probability.
• Fixing bug in logistic regression local predictions with operating points.

4.13.7 (2018-01-02)

• Changing local predictions with threshold to meet changes in backend.
• Adding support for configurations REST API calls.

4.13.6 (2017-12-05)

• Fixing predict confidence method in local ensembles.

4.13.5 (2017-11-23)

• Adding operating point local predictions to deepnets.

4.13.4 (2017-11-21)

• Fixing bug in local ensemble predictions with operating points.
• Fixing bug for local EnsemblePredictor class.

4.13.3 (2017-11-14)

• Fixing bug in local ensemble predictions for inputs that don't match the expected field types.

4.13.2 (2017-11-14)

• Adding left out static files for local ensemble predictor functions.

4.13.1 (2017-11-10)

• Refactoring local BoostedTrees and adding the EnsemblePredictor to use the local predict functions of each model to generate the ensemble prediction.

4.13.0 (2017-11-07)

• Adding operating point thresholds to local model, ensemble and logistic regression predictions.

4.12.1 (2017-10-12)

• Fixing bug in the local Deepnet predictions when numpy is not installed.

4.12.0 (2017-10-04)

• Adding support for Deepnets REST API calls and local predictions using the local Deepnet object.

4.11.3 (2017-09-29)

• Fixing bug in the local Ensemble object. Failed to use the stored ensemble object.

4.11.2 (2017-07-29)

• Fixing bug in source uploads using Python3 when reading data from stdin.

4.11.1 (2017-06-23)

• Fixing bug in source uploads using Python3 when a category is set.

4.11.0 (2017-06-23)

• Adding REST methods for managing time-series and local time-series object to create forecasts.

4.10.5 (2017-07-13)

• Fixing bug in the sources upload using Python3. Server changes need the content-type of the file to be sent.

4.10.4 (2017-06-21)

• Fixing bug in the local model predicted distributions for weighted models.
• Fixing bug in predicted probability for local model predictions using weighted models.

4.10.3 (2017-06-07)

• Changing boosted local ensembles predictions to match the improvements in API.
• Fixing bug in association rules export to CSV and lisp for rules with numeric attributes.

4.10.2 (2017-05-23)

• Fixing bug: local Model object failed when retrieving old JSON models from local storage.

4.10.1 (2017-05-15)

• Internal refactoring preparing for extensions in BigMLer.

4.10.0 (2017-05-05)

• Adding predic_probability and predict_confidence methods to local model and ensemble.
• Internal refactoring of local model classes preparing for extensions in BigMLer.

4.9.2 (2017-03-26)

• Fixing bug: local model slugifying fails when fields have empty names.

4.9.1 (2017-03-23)

• Adding methods to local cluster: closest data points from a reference point and centroids ordered from a reference point.
• Modifying internal codes in MultiVote class.

4.9.0 (2017-03-21)

• Adding boosted ensembles to the local Ensemble object.

4.8.3 (2017-03-01)

• Fixing bug in local logistic regression predictions when a constant field is forced as input field.

4.8.2 (2017-02-09)

• Fixing bug: Adapting to changes in Python 3.6 which cause the connection to the API using SSL to fail.

4.8.1 (2017-01-11)

• Changing local association parameters to adapt to API docs specifications.

4.8.0 (2017-01-08)

• Adapting to final format of local association sets and adding tests.

4.7.3 (2016-12-03)

• Bug fixing: query string is allowed also for project get calls.

4.7.2 (2016-12-02)

• Allowing a query string to be added to get calls for all the resource types.

4.7.1 (2016-12-01)

• Improving the Fields object: extracting fields structure from topic models.
• Bug fixing: Local Topic Distributions failed when tokenizing inputs with sequences of separators.

4.7.0 (2016-11-30)

• Adding REST methods for the new resource types: Topic Model, Topic Distribution, Batch Topic Distribution.
• Adding local Topic Model object.

4.6.10 (2016-10-26)

• Improving local cluster object to fill in missing numerics for clusters with default numeric values.

4.6.9 (2016-09-27)

• Fixing bug in tests for anomaly detector and ill-formatted comments.
• Adapting tests to new logistic regression default value for balance_fields.

4.6.8 (2016-09-22)

• Adding optional information to local predictions.
• Improving casting for booleans in local predictions.
• Improving the retrieval of stored or remote resources in local predictor objects.

4.6.7 (2016-09-15)

• Changing the type for the bias attribute to create logistic regressions to boolean.

4.6.6 (2016-08-02)

• Improving message for unauthorized API calls adding information about the current domain.

4.6.5 (2016-07-16)

• Fixing bug in local model. Fixing predictions for weighted models.

4.6.4 (2016-07-06)

• Fixing bug in delete_execution method. The delete call now has a query_string.

4.6.3 (2016-06-25)

• Fixing bug in local logistic regression predictions' format.

4.6.2 (2016-06-20)

• Adding local logistic regression as argument for evaluations.

4.6.1 (2016-06-12)

• Adapting local logistic regression object to new coefficients format and adding field_codings attribute.

4.6.0 (2016-05-19)

• Adding REST methods to manage new types of whizzml resources: scripts, executions and libraries.
• Fixing bug in logistic regression predictions for datases with text fields. When input data has only one term and all token mode is used, local and remote predictions didn't match.

4.5.3 (2016-05-04)

• Improving the cluster report information.
• Fixing bug in logistic regression predictions. Results differred from the backend predictions when date-time fields were present.

4.5.2 (2016-03-24)

• Fixing bug in model's local predictions. When the model uses text fields and the field contents are missing in the input data, the prediction does not return the last prediction and stop. It now follows the "does not contain" branch.

4.5.1 (2016-03-12)

• Adding method to Fields object to produce CSV summary files.
• Adding method to Fields object to import changes in updatable attributes from CSV files or strings.

4.5.0 (2016-02-08)

• Adapting association object to the new syntax of missing values.
• Improving docs and comments for the proportional strategy in predictions.
• Fixing bug: centroid input data datetime fields are optional.

4.4.2 (2016-01-06)

• Adapting logistic regression local object to the new missing_numeric parameter.

4.4.1 (2015-12-18)

• Fixing bug: summarized path output failed when adding missing operators.

4.4.0 (2015-12-15)

• Adding REST API calls for association rules and local Association object.
• Adapting local model, cluster, anomaly and logistic regression objects to new field type: items.
• Fixing bug: wrong value of giny impurity
• Fixing bug: local model summary failed occasionally when missings were used in a numeric predicate.
• Fixing bug: wrong syntax in flatline filter method of the tree object.

4.3.4 (2015-12-10)

• Fixing bug: Logistic regression object failed to build when using input fields or non-preferred fields in dataset.

4.3.3 (2015-11-30)

• Fixing bug: Anomaly object failed to generate the filter for new datasets when text empty values were found.

4.3.2 (2015-11-24)

• Adding verify and protocol options to the existing Domain class constructor to handle special installs.

4.3.1 (2015-11-07)

• Fixing bug: Local logistic regression predictions differ when input data has contents in a text field but the terms involved do not appear in the bag of words.

4.3.0 (2015-10-16)

• Adding logistic regression as a new prediction model.

4.2.2 (2015-10-14)

• Fixing bug: Fields object failed to store the correct objective id when the objective was in the first column.

4.2.1 (2015-10-14)

• Fixing bug: Improving error handling in download_dataset method.

4.2.0 (2015-07-27)

• Adding REST methods to manage new type of resource: correlations.
• Adding REST methods to manage new type of resource: tests.
• Adding min and max values predictions for regression models and ensembles.
• Fixing bug: Fields object was not retrieving objective id from the resource info.

4.1.7 (2015-08-15)

• Fixing bug: console messages failed when used with Python3 on Windows.

4.1.6 (2015-06-25)

• Fixing bug: Removing id fields from the filter to select the anomalies listed in the Anomaly object from the origin dataset.

4.1.5 (2015-06-06)

• Fixing bug: create_source method failed when unicode literals were used in args.

4.1.4 (2015-05-27)

• Ensuring unique ordering in MultiVote categorical combinations (only needed in Python 3).

4.1.3 (2015-05-19)

• Adapting code to handle uploading from String objects.
• Adding models creation new origin resources: clusters and centroids.

4.1.2 (2015-04-28)

• Fixing bug in summarize method for local models. Ensuring unicode use and adding tests for generated outputs.

4.1.1 (2015-04-26)

• Fixing bug in method to print the fields in the anomaly trees.
• Fixing bug in the create_source method for Python3. Creation failed when the tags argument was used.

4.1.0 (2015-04-14)

• Adding median based predictions to ensembles.

4.0.2 (2015-04-12)

• Fixing bug: multimodels median predictions failed.

4.0.1 (2015-04-10)

• Adding support for median-based predictions in MultiModels.

4.0.0 (2015-04-10)

• Python 3 added to supported Python versions.
• Test suite migrated to nose.

3.0.3 (2015-04-08)

• Changing setup to ensure compatible Python and requests versions.
• Hiding warnings when SSL verification is disabled.

3.0.2 (2015-03-26)

• Adding samples as Fields generator resources

3.0.1 (2015-03-17)

• Changing the Ensemble object init method to use the max_models argument also when loading the ensemble fields to trigger garbage collection.

3.0.0 (2015-03-04)

• Adding Google App Engine support for remote REST calls.
• Adding cache_get argument to Ensemble constructor to allow getting local model objects from cache.

2.2.0 (2015-02-26)

• Adding lists of local models as argument for the local ensemble constructor.

2.1.0 (2015-02-22)

• Adding distribution and median to ensembles' predictions output.

2.0.0 (2015-02-12)

• Adding REST API calls for samples.

1.10.8 (2015-02-10)

• Adding distribution units to the predict method output of the local model.

1.10.7 (2015-02-07)

• Extending the predict method in local models to get multiple predictions.
• Changing the local model object to add the units used in the distribution and the add_median argument in the predict method.

1.10.6 (2015-02-06)

• Adding the median as prediction for the local model object.

1.10.5 (2014-01-29)

• Fixing bug: centroids failed when predicted from local clusters with summary fields.

1.10.4 (2014-01-17)

• Improvements in docs presentation and content.
• Adding tree_CSV method to local model to output the nodes information in CSV format.

1.10.3 (2014-01-16)

• Fixing bug: local ensembles were not retrieved from the stored JSON file.
• Adding the ability to construct local ensembles from any existing JSON file describing an ensemble structure.

1.10.2 (2014-01-15)

• Source creation from inline data.

1.10.1 (2014-12-29)

• Fixing bug: source upload failed in old Python versions.

1.10.0 (2014-12-29)

• Refactoring the BigML class before adding the new project resource.
• Changing the ok and check_resource methods to download lighter resources.
• Fixing bug: cluster summarize for 1-centroid clusters.
• Fixing bug: adapting to new SSL verification in Python 2.7.9.

1.9.8 (2014-12-01)

• Adding impurity to Model leaves, and a new method to select impure leaves.
• Fixing bug: the Model, Cluster and Anomaly objects had no resource_id attribute when built from a local resource JSON structure.

1.9.7 (2014-11-24)

• Adding method in Anomaly object to build the filter to exclude anomalies from the original dataset.
• Basic code refactorization for initial resources structure.

1.9.6 (2014-11-09)

• Adding BIGML_PROTOCOL, BIGML_SSL_VERIFY and BIGML_PREDICTION_SSL_VERIFY environment variables to change the default corresponding values in customized private environments.

1.9.5 (2014-11-03)

• Fixing bug: summarize method breaks for clusters with text fields.

1.9.4 (2014-10-27)

• Changing MultiModel class to return in-memory list of predictions.

1.9.3 (2014-10-23)

• Improving Fields and including the new Cluster and Anomalies fields structures as fields resources.
• Improving ModelFields to filter missing values from input data.
• Forcing garbage collection in local ensemble to lower memory usage.

1.9.2 (2014-10-13)

• Changing some Fields exceptions handling.
• Refactoring api code to handle create, update and delete methods dynamically.
• Adding connection info string for printing.
• Improving tests information.

1.9.1 (2014-10-10)

• Adding the summarize and statistics_CSV methods to local cluster object.

1.9.0 (2014-10-02)

• Adding the batch anomaly score REST API calls.

1.8.0 (2014-09-09)

• Adding the anomaly detector and anomaly score REST API calls.
• Adding the local anomaly detector.

1.7.0 (2014-08-29)

• Adding to local model predictions the ability to use the new missing-combined operators.

1.6.7 (2014-08-05)

• Fixing bug in corner case of model predictions using proportional missing strategy.
• Adding the unique path to the first missing split to the predictions using proportional missing strategy.

1.6.6 (2014-07-31)

• Improving the locale handling to avoid problems when logging to console under Windows.

1.6.5 (2014-07-26)

• Adding stats method to Fields to show fields statistics.
• Adding api method to create a source from a batch prediction.

1.6.4 (2014-07-25)

• Changing the create methods to check if origin resources are finished by downloading no fields information.

1.6.3 (2014-07-24)

• Changing some variable names in the predict method (add_count, add_path) and the prediction structure to follow other bindigns naming.

1.6.2 (2014-07-19)

• Building local model from a JSON model file.
• Predictions output can contain confidence, distribution, instances and/or rules.

1.6.1 (2014-07-09)

• Fixing bug: download_dataset method did not return content when no filename was provided.

1.6.0 (2014-07-03)

• Fixing bug: check valid parameter in distribution merge function.
• Adding downlod_dataset method to api to export datasets to CSV.

1.5.1 (2014-06-13)

• Fixing bug: local clusters' centroid method crashes when text or categorical fields are not present in input data.

1.5.0 (2014-06-05)

• Adding local cluster to produce centroid predictions locally.

1.4.4 (2014-05-23)

• Adding shared urls to datasets.
• Fixing bug: error renaming variables.

1.4.3 (2014-05-22)

• Adding the ability to change the remote server domain in the API connection constructor (for VPCs).
• Adding the ability to generate datasets from clusters.

1.4.2 (2014-05-20)

• Fixing bug when using api.ok method for centroids and batch centroids.

1.4.1 (2014-05-19)

• Docs and test updates.

1.4.0 (2014-05-14)

• Adding REST methods to manage clusters, centroids and batch centroids.

1.3.1 (2014-05-06)

• Adding the average_confidence method to local models.
• Fixing bug in pprint for predictions with input data keyed by field names.

1.3.0 (2014-04-07)

• Changing Fields object constructor to accept also source, dataset or model resources.

1.2.2 (2014-04-01)

• Changing error message when create_source calls result in http errors to standarize them.
• Simplifying create_prediction calls because now API accepts field names as input_data keys.
• Adding missing_counts and error_counts to report the missing values and error counts per field in the dataset.

1.2.1 (2014-03-19)

• Adding error to regression local predictions using proportional missing strategy.

1.2.0 (2014-03-07)

• Adding proportional missing strategy to MultiModel and solving tie breaks in remote predictions.
• Adding new output options to model's python, rules and tableau outputs: ability to extract the branch of the model leading to a certain node with or without the hanging subtree.
• Adding HTTP_TOO_MANY_REQUESTS error handling in REST API calls.

1.1.0 (2014-02-10)

• Adding Tableau-ready ouput to local model code generators.

1.0.6 (2014-02-03)

• Fixing getters: getter for batch predictions was missing.

1.0.5 (2014-01-22)

• Improving BaseModel and Model. If they receive a partial model structure with a correct model id, the needed model resource is downloaded and stored (if storage is enabled in the given api connection).
• Improving local ensemble. Adding a new fields attribute that contains all the fields used in its models.

1.0.4 (2014-01-21)

• Adding a summarize method to local ensembles with data distribution and field importance information.

1.0.3 (2014-01-21)

• Fixes bug in regressions predictions with ensembles and plurality without confidence information. Predictions values were not normalized.
• Updating copyright information.

1.0.2 (2014-01-20)

• Fixes bug in create calls: the user provided args dictionaries were updated inside the calls.

1.0.1 (2014-01-05)

• Changing the source for ensemble field importance computations.
• Fixes bug in http_ok adding the valid state for updates.

1.0.0 (2013-12-09)

• Adding more info to error messages in REST methods.
• Adding new missing fields strategy in predict method.
• Fixes bug in shared models: credentials where not properly set.
• Adding batch predictions REST methods.

0.10.3 (2013-12-19)

• Fixes bug in local ensembles with more than 200 fields.

0.10.2 (2013-12-02)

• Fixes bug in summarize method of local models: field importance report crashed.
• Fixes bug in status method of the BigML connection object: status for async uploads of source files crashed while uploading.

0.10.1 (2013-11-25)

• Adding threshold combiner to MultiModel objects.

0.10.0 (2013-11-21)

• Adding a function printing field importance to ensembles.
• Changing Model to add a lightweight BaseModel class with no Tree information.
• Adding function to get resource type from resource id or structure.
• Adding resource type checks to REST functions.
• Adding threshold as new combination method for local ensembles.

0.9.1 (2013-10-17)

• Fixes duplication changing field names in local model if they are not unique.

0.9.0 (2013-10-08)

• Adds the environment variables and adapts the create_prediction method to create predictions using a different prediction server.
• Support for shared models.

0.8.0 (2013-08-10)

• Adds text analysis local predict function
• Modifies outputs for text analysis: rules, summary, python, hadoop

0.7.5 (2013-08-22)

• Fixes temporarily problems in predictions for regression models and ensembles
• Adds en-gb to the list of available locales, avoiding spurious warnings

0.7.4 (2013-08-17)

• Changes warning logger level to info

0.7.3 (2013-08-09)

• Adds fields method to retrieve only preferred fields
• Fixes error message when no valid resource id is provided in check_resource

0.7.2 (2013-07-04)

• Fixes check_resource method that was not using query-string data
• Add list of models as argument in Ensemble constructor
• MultiModel has BigML connection as a new optional argument

0.7.1 (2013-06-19)

• Fixes Multimodel list_models method
• Fixes check_resource method for predictions
• Adds local configuration environment variable BIGML_DOMAIN replacing BIGML_URL and BIGML_DEV_URL
• Refactors Ensemble and Model's predict method

0.7.0 (2013-05-01)

• Adds splits in datasets to generate new datasets
• Adds evaluations for ensembles

0.6.0 (2013-04-27)

• REST API methods for model ensembles
• New method returning the leaves of tree models
• Improved error handling in GET methods

0.5.2 (2013-03-03)

• Adds combined confidence to combined predictions
• Fixes get_status for resources that have no status info
• Fixes bug: public datasets, that should be downloadable, weren't

0.5.1 (2013-02-12)

• Fixes bug: no status info in public models, now shows FINISHED status code
• Adds more file-like objects (e.g. stdin) support in create_source input
• Refactoring Fields pair method and Model predict method to increase
• Adds some more locale aliases

0.5.0 (2013-01-16)

• Adds evaluation api functions
• New prediction combination method: probability weighted
• Refactors MultiModels lists of predictions into MultiVote
• Multimodels partial predictions: new format

0.4.8 (2012-12-21)

• Improved locale management
• Adds new features to MultiModel to allow local batch predictions
• Improved combined predictions
• Adds local predictions options: plurality, confidence weighted

0.4.7 (2012-12-06)

• Warning message to inform of locale default if verbose mode

0.4.6 (2012-12-06)

• Fix locale code for windows

0.4.5 (2012-12-05)

• Fix remote predictions for input data containing fields not included in rules

0.4.4 (2012-12-02)

• Tiny fixes
• Fix local predictions for input data containing fields not included in rules
• Overall clean up

0.4.3 (2012-11-07)

• A few tiny fixes
• Multi models to generate predictions from multiple local models
• Adds hadoop-python code generation to create local predictions

0.4.2 (2012-09-19)

• Fix Python generation
• Add a debug flag to log https requests and responses
• Type conversion in fields pairing

0.4.1 (2012-09-17)

• Fix missing distribution field in new models
• Add new Field class to deal with BigML auto-generated ids
• Add by_name flag to predict methods to avoid reverse name lookups
• Add summarize method in models to generate class grouped printed output

0.4.0 (2012-08-20)

• Development Mode
• Remote Sources
• Bigger files streamed with Poster
• Asynchronous Uploading
• Local Models
• Local Predictions
• Rule Generation
• Python Generation
• Overall clean up

0.3.1 (2012-07-05)

• Initial release for the "andromeda" version of BigML.io.