A Google+ Interface for R
plusser provides an API interface to Google+ so that posts,
profiles and pages can be automatically retrieved.
The central aim of plusser is to facilitate access to the Google+ API for
retrieving posts and profiles from Google's social network. This package is
especially suited to harvest texts and covariate data for social media analysis.
For social media data harvests, this package provides three key functions:
harvestPage to retrieve messages a user posted publicly
harvestProfile to retrieve data from a user's profile (i.e. about the user)
harvestActivity to retrieve the +1ers and resharers of a specific post.
The results are intentionally kept simple for now, so that they can be easily
used in text mining or predictive modeling contexts. There are basic parsing
functions provided. Users can also use their own parsing functions if required.
At the moment, this package
implements basic authentication (via API keys) only. Full OAuth 2.0
authentication will be implemented later. This will then also be the basis for
functions posting in a user's name.
Changes in version 0.4-0 (2014-04-27)
- parseAttachment() can be used to extract the attachments from posts.
Dependent on the type of attachment, different fields are being extracted.
Owing to its modular design, it is possible to extend the parsing of future attachment types.
- harvestPage() and harvestProfile() now use pluggable parsing functions. All
parsing code has been put into external functions like parsePost() and parseProfile().
The harvest*() functions can be called with these or user-contributed parsing functions.
Thanks go to Bud Gibson for the idea.
- parsePost() is not returning the number of attachments (nA) anymore, as it was always 0 or 1.
Rather, it now returns the type of attachment a post has (or NA, if there is no attachment).
Changes in version 0.3-6 (2014-04-22)
- harvestPage had incorrectly ignored the nextPage token if called recursively.
Thanks to Bud Gibson for pointing this out.
Changes in version 0.3-5 (2014-04-01)
- Additional person fields removed from DESCRIPTION.
Changes in version 0.3-4 (2014-03-24)
- DESCRIPTION now has a Maintainer field, as binary build servers require this.
Changes in version 0.3-3 (2014-03-14)
- Harvesting profiles now returns also a profile's website (if it has one).
- Searching for profiles now returns an additional column differentiating
between profiles and pages.
- Some columns of the results when harvesting profiles were always NA because
the corresponding field had not been selected. All fields are now selected.
- Also when harvesting profiles, type was always NA; this is now fixed.
Changes in version 0.3-2 (2014-03-05)
- Dropping the knitr dependency for now, as a vignette is unlikely to be
included due to authentication issues.
- The Google+ API key is now stored in a private environment instead of
.Globalenv. This is required to pass CRAN checks and safeguards against
overwriting userspace variables.
Changes in version 0.3-1 (2014-03-04)
- The age of a post (as returned by 'parsePost') is now always in days.
Previously, posts that were younger than a day had their age returned in
hours. This led to inconsistencies in the interpretation of age.
Changes in version 0.3-0 (2014-03-03)
- harvestProfile is now vectorized, accepting a vector of user IDs.
- All functions that depend on it now fail gracefully if no API key has been
- Functions now use a basic heuristic in validating their respective arguments.
Changes in version 0.2-0 (2014-02-28)
- Gained two functions to harvest user profiles and user IDs of users acting on
- Gained a function to search profiles for query strings.
- Gained a function to search for a query in posts.
- More elaborate package documentation.
- Every function now contains at least on example.
- Retrieval of paginated results oriented on target values.
- All user supplied strings are now URL encoded (escaped).
Changes in version 0.1-0 (2014-02-26)
- Initial rudimentary package version.