NEWS.md

In facebookresearch/Radlibrary: Facebook Ad Library API Tool

Radlibrary 0.5.0

• Added support for new fields available for ads delivered to the EU
• age_country_gender_reach_breakdown: Demographic distribution of ads delivered to the EU by country, gender, and age
• beneficiary_payers: Reported beneficiaries and payers for the ad
• eu_total_reach: The estimated combined ad reach for all locations inside the European Union
• target_ages: The age ranges selected for ad targeting in the EU.
• target_gender: The genders selected for ad targeting in the EU.
• target_locations: The locations included or excluded for ad targeting in the EU.
• Added support for controlling whether to include ads that are removed for community standards violations (unmask_removed_content)
• Various small bug fixes and updates and more tests

Radlibrary 0.4.3

• Very minor bug fix to make R CMD Check pass
• there has been a line that was checking whether something was an instance of a class in the wrong way

Radlibrary 0.4.2

• Fixed an issue where searches by page_id were returning fewer than expected results.

Radlibrary 0.4.1

• Fix some documentation issues
• Update the nested columns vignette

Radlibrary 0.4.0

New Fields

The Ad Library API provides more data than it used to, and now Radlibrary can handle it. All of the current fields are supported, which at this time include:

• id
• ad_creation_time
• ad_creative_bodies
• ad_creative_link_captions
• ad_creative_link_descriptions
• ad_creative_link_titles
• ad_delivery_start_time
• ad_delivery_stop_time
• ad_snapshot_url
• bylines
• currency
• estimated_audience_size
• impressions
• languages
• page_id
• page_name
• publisher_platforms
• spend
• demographic_distribution
• delivery_by_region

Nested Columns

The Ad Library API has evolved to return richer data for some fields which may be multi-valued. For instance, the publisher_platform field can contain multiple values: Facebook, Instagram, etc. Converting responses to a tibble now makes these into list columns.

Similarly, some values like demographic_distribution and delivery_by_region return very rich data. Previously, it was necessary to request this data separately as a geo_table or demographic_table. Now these values are converted into nested tibble columns. See vignette("nested_columns") for more.

Other small-to-medium changes

• Ads now have a unique identifier simply called id. This is now the favored unique identifier rather than ad_snapshot_url, which was previously a required field in all calls to the API.
• Converting a data response to tibble will no longer create a whole bunch of unused columns for fields that weren't requested. Only the fields that were requested are included now.

Radlibrary 0.3.0

• Added ability to filter for potential reach
• Added date filters
• Removed some deprecated fields
• Bug fixes

Radlibrary 0.2.5

• Fixed an issue where ad tables couldn't be created when spend and impressions were not requested
• Require ad_snapshot_url to be present in all queries

Radlibrary 0.2.4

• Fixed getPass bug in Rstudio server
• Only warn once about environment variables on Linux
• fix typos in docs

Radlibrary 0.2.3

• improve docs
• pass rhub windows build check

Radlibrary 0.2.2

• bug fixes
• table types vignette

Radlibrary 0.2.1

• tests and bugfixes
• improved documentation

Radlibrary 0.2.0

• CI is set up and passes
• API Requests set the user agent to Radlibrary
• Access tokens are censored in URLs by default
• Spend and Impressions columns have numeric upper and lower bounds
• Documentation
• Bug fixes

Radlibrary 0.1.0

• Secrets are stored securely in keychain
• Requests return data_objects that can be converted to tibbles
• Paginated API requests built
• A bunch of documentation

Radlibrary 0.0.0.9001

• Added a NEWS.md file to track changes to the package.

facebookresearch/Radlibrary documentation built on Feb. 2, 2024, 12:01 a.m.