v0.2.13

Fixes:

• Fixes a bug that lead to a memory leak on the AzureML web service during consume(). Load exportenv only once (during first call) #117

v0.2.12 CRAN release

This version was released to CRAN on 2017-07-12

Fixes:

• Upload packages from a local repository using publishWebservice() #109

Enhancements

• Produce more informative error messages from consume() (#57)
• Better documentation and examples for endpoint settings, especially for regional AML instances (#105)

This version also contains many other internal improvements that probably won't be visible to most users

v0.2.11 Bug fix and refactor release

This release fixes multiple internal issues:

• Add additional skip logic to skip tests on CRAN and if no Internet connection tests (#114)
• Fix unit tests and code for download.datasets() to deal with multiple datasets bug tests (#111)
• Upload packages from a local repository using publishWebservice() enhancement (#109)
• Missing workspace parameter on download.datasets() leads to cryptic error message bug (#93)
• Fix bug where example for download.datasets() doesn't work (#104)

v0.2.10 CRAN Release

This version was published on CRAN on 2016-02-01

Functional changes:

• None

Other changes:

• Added images to the workspace() function help in the PDF manual
• Ensured only a minimal subset of tests run on CRAN
• Replaced vignette with text only, to ensure no functionality runs on CRAN

v0.2.9 Bug fix release

This release fixes a bug that was inadvertently introduced after v0.2.5. Specifically, users reported errors when calling the workspace() function when they had no ~/.azureml/settings.json file.

The error message was:

Error in file.exists(config): invalid 'file' argument
Traceback:

1. workspace()
2. validate.AzureML.config(config, stopOnError = TRUE)
3. file.exists(config)

Bug fixes:

• workspace() throws error if config file not found #83
• Service does not authenticate with europewest API endpoint #89
• AzureML.config option not consistently set #86

Enhancements:

• AzureML.config option not consistently set #86

Other changes:

• Updated copyright statements to (c) 2015-2016 Microsoft Corporation

v0.2.8 Usability improvement release

Enhancements:

• Automatically calls datasets() when a workspace is created to validate the workspace id and auth token
• More informative error messages if AzureML responds with an error message
• Added images in the HTML help for workspace() to show where to find id and auth

Behind the scenes:

• Deals with differences in how the PROD and INT versions of the AzureML service returns results
• Refactored calls to always use expontential backoff - this means that all calls to AzureML retry several times if the service returns an error, and the elapsed time between calls grow exponentially.

v0.2.7 Bug fix release

This release fixes a bug that required you to have a settings file at ~/.azureml/settings.json.

The setttings file is optional - if it does not exist, the user has to specify the workspace id and authorization code, but the other api settings are taken from defaults.

v0.2.6 Bug fix release

Enhancements:

• Change default values for exponential backoff. Now the default time delay for retry is 1 second (down from 10s), but the time delay grows exponentially (1, 3, 7, 15, ...) and number of retries is increased to 6 (up from 3).

Bug fixes

• Use consistent api_endpoint URLs for communication with the AzureML service (#67)
• Fix some errors in reading the ~/.azureml/settings.json file, and setting the defaults correctly if not found in the settings file (#65)

v0.2.5 Bug fix release

Enhancements:

• Read the api_endpoint and management_endpoint from the config file (settings.json), if provided.

Bug fixes

• When the user attempts to upload a dataset with duplicate name, intercept and provide a meaningful error.

v0.2.4 Bug fix release

Bug fixes:

• Use double quotes as the quote character for downloading datasets. This fixes a bug with data sets that contain embedded single quotes
• Fix a number of issues with updateWebservice(). For example, no longer require a name argument to be specified.

Improvements

• Much expanded examples for publishing a function as a webservice, both in the help for publishWebservice() and the vignette
• More descriptive error messages for consume()
• In updateWebservice(), automatically determine the ServiceId, if available in the supplied endpoint.
• Improved documentation for download.intermediate.dataset()