NEWS.md

sentometrics 1.0.0

• Version bump associated with publication of vignette in the Journal of Statistical Software.

sentometrics 0.8.4

• Alignment with released quanteda v3.0.

sentometrics 0.8.3

• Features (or docvars) with names "id", "sentence_id", "date", "word_count" or "texts" will not be accepted even when numeric, to avoid duplicate column names down the line. A clear error message is issued to alert users.
• Replacement of order() calls on data.frames where needed to avoid CRAN complaints.
• Some small documentation fixes.

sentometrics 0.8.2

• Some documentation fixes.
• Release of a pkgdown website.
• Fixed bug in sento_corpus() function that did not always order input correctly by date.
• Fixed two minor bugs in summary.sento_measures(); the first one prevented printing of document-level weighting schemes, the second one did not remove NAs when averaging over correlations.
• Small bug fix in yearly aggregation (it did not account for the fact that 1970-01-01 is considered day zero).
• Dropped horizontal 0-line automatically added in the plot.sento_measures() function as it distorts graphs of time series with values far away from zero.
• Stopped exporting all defunct functions to clean up namespace.
• The function print.sento_corpus() now shows when corpus is multi-lingual.

sentometrics 0.8.1

• Alignment with released quanteda v2.0.

sentometrics 0.8.0

• New function: print.sento_corpus().
• Package update followed by release of a substantial update of the vignette (see https://doi.org/10.2139/ssrn.3067734).
• Changed some warning() calls to message() calls to be more kind to the user.
• Altered internal code to comply with the corpus object from quanteda >= v2.0.
• Dropped all "TF"-inspired weights for within-document aggregation except for "TFIDF", and made this option return the same sentiment scores as would when using the quanteda package (see the example on https://sentometrics-research.com/sentometrics/articles/examples/sentiment.html).

sentometrics 0.7.6

• Fixed memory allocation issue in the compute_sentiment() function.

sentometrics 0.7.5

• New functions: as.data.table.sento_corpus(), as.data.frame.sento_corpus(), and as.data.frame.sento_measures().
• Embedded a small workaround in plot.attributions() to guaranty same plotting behaviour after update of ggplot2 package that gave buggy output for the geom_area() layer.
• Integrated for overall consistency measures_global() into the aggregate.sento_measures() function, adding a do.global argument to enact it.
• Slightly changed the clusters-based sentence-level sentiment computation (different weighting of adversative conjunctions).
• Clarified the documentation for the peakdates() and peakdocs() functions.
• Put the Shiny application made available in previous package update (i.e., the sento_app() function) in a separate sole-purpose package sentometrics.app (see https://github.com/sborms/sentometrics.app).
• Moved the data.table package from Depends to Imports (see Rdatatable/data.table#3076).
• No change by reference of input sentiment objects in the merge.sentiment() function anymore, and modified the merging to give for instance a simple column binding of sentment methods when all else is equal.
• Correct pass-through of default how argument in the compute_sentiment() function.
• Added a few adversative conjunctions to all word lists in list_valence_shifters.
• Added a do.normalize option to the weights_beta() and weights_exponential() functions.
• Added a do.inverse option to the weights_exponential() function and associated do.inverseExp argument in the ctr_agg() function.
• Modified some names of options for within-document or within-sentence aggregation (i.e., across tokens): "squareRootCounts" into "proportionalSquareRoot", "invertedExponential" into "inverseExponential", and "invertedUShaped" into "inverseUShaped".
• Corrected the numerator (number of documents or sentences instead of token frequency) in all weighting schemes involving the inverse document frequency (IDF).
• Aligned all formulas concerning the exponential weighting curves.
• The compute_sentiment() function now also can do a sentence-level calculation using the bigrams valence shifting approach.
• Fixed a small bug that did not allow to have different valence shifters lists for a multi-language sentiment calculation.

sentometrics 0.7.0

• New functions: measures_update(), subset.sento_measures(), as.sentiment(), as.sento_measures(), as.data.table.sentiment(), corpus_summarize(), sento_app(), and aggregate.sento_measures().
• Defunct all deprecated functions as well as the functions replaced by the new functions (wiping the slate clean...).
• Handled reverse dependency issue raised by quanteda developers regarding their new corpus object.
• Renamed the class objects coming from any sento_xyz() function into the name of the function (e.g., the sento_measures() function now gives a sento_measures object instead of a sentomeasures object).
• Fixed a small bug in the aggregate.sento_measures() (previously measures_merge()) function to take the mean instead of the sum in a particular case.
• Added many more within- and across-document weighting schemes (see the get_hows() function for an overview).
• Added the flexibility to do an explicit sentence-by-sentence sentiment computation (see do.sentence argument in the compute_sentiment() function).
• Added the flexibility to create a multi-language sento_corpus object to do a multi-language sentiment computation (applying different lexicons to texts written in different languages).
• Expanded the compute_sentiment() function to also take tm SimpleCorpus and VCorpus objects.
• Added the tm and NLP packages to Suggests.

sentometrics 0.5.6

• New function: peakdates().
• Modified the purpose of the peakdocs() function and added a peakdates() function to properly handle the entire functionality of extracting peaks.
• A series of documentation fixes.

sentometrics 0.5.5

• New functions: sentiment_bind(), and to_sentiment().
• Defined replacement (of lexicons and names) for a sentolexicons object.
• Properly handled lag = 1 in the ctr_agg() function, and set weights to 1 by default for n = 1 in the weights_beta() function.
• Solved single failing test for older R version (3.4.4).
• Removed the abind package from Imports.
• Removed the zoo package from Imports, by replacing the single occurrence of the zoo::na.locf() function by the fill_NAs() helper function (written in Rcpp).
• Extended the quanteda::docvars() replacement method to a sentocorpus object.
• Modified information criterion estimators for edge cases to avoid them turning negative.
• Dropped the "x" output element from a sentomodel object (for large samples, this became too memory consuming).
• Dropped the "howWithin" output element from a sentomeasures object, and simplified a sentiment object into a data.table directly instead of a list.
• Expanded the do.shrinkage.x argument in the ctr_model() function to a vector argument.
• Added a do.lags argument to the attributions() function, to be able to circumvent the most time-consuming part of the computation .
• Imposed a check in the sento_measures() function on the uniqueness of the names within and across the lexicons, features and time weighting schemes.
• Solved a bug in the measures_merge() function that made full merging not possible.
• The n argument in the peakdocs() function can now also be specified as a quantile.

sentometrics 0.5.1

• Minor modifications to resolve few CRAN issues.
• Set default value of nCore argument in the compute_sentiment() and ctr_agg() functions to 1.
• Classed the output of the compute_sentiment.sentocorpus() function as a sentiment object, and modified the aggregate() function to aggregate.sentiment().

sentometrics 0.5.0

• New functions: weights_beta(), get_dates(), get_dimensions(), get_measures(), and get_loss_data().
• Renamed following functions: to_global() to measures_global(), perform_agg() to aggregate(), almons() to weights_almon(), exponentials() to weights_exponential(), setup_lexicons() to sento_lexicons(), retrieve_attributions() to attributions(), plot_attributions() to plot.attributions().
• Defunct the ctr_merge() function, so that all merge parameters have to be passed on directly to the measures_merge() function.
• Expanded the use of the center and scale arguments in the scale() function.
• Added the dateBefore and dateAfter arguments to the measures_fill() function, and dropped NA option of its fill argument.
• Added a "beta" time aggregation option (see associated weights_beta() function).
• Corrected update of "attribWeights" element of output sentomeasures object in required measures_xyz() functions.
• Added a new attribution dimension ("lags") to the attributions() function, and corrected some edge cases.
• Made a slight correction to the information criterion estimators.
• Added a lambdas argument to the ctr_model() function, directly passed on to the glmnet::glmnet() function if used.
• Omitted do.combine argument in measures_delete() and measures_select() functions to simplify.
• Expanded set of unit tests, included a coverage badge, and added covr to Suggests.
• Reimplementation (and improved documentation) of the sentiment calculation in the compute_sentiment() function, by writing part of the code in Rcpp relying on RcppParallel (added to Imports); there are now three approaches to computing sentiment (unigrams, bigrams and clusters).
• Replaced the dfm argument in the compute_sentiment() and ctr_agg() functions by a tokens. argument, and altered the input and behaviour of the nCore argument in these same two functions.
• Switched from the quanteda package to the stringi package for more direct tokenization.
• Trimmed the list_lexicons and list_valence_shifters built-in word lists by keeping only unigrams, and included same trimming procedure in the sento_lexicons() function.
• Added a column type "t" to the list_valence_shifters built-in word list, and reset values of the "y" column from 2 to 1.8 and from 0.5 to 0.2.
• Updated the epu built-in dataset with the newest available series, up to July 2018.
• Corrected the word 'sparesly' to 'sparsely' in list_valence_shifters[["en"]].
• Further shortened project page to the bare essence.
• Omitted statement printed ('Compute sentiment... Done.') in the compute_sentiment() function.
• Slightly modified print() generic for a sentomeasures object.
• Dropped the "tf-idf" option for within-document aggregation in the ctr_agg() function.
• The sento_lexicons() function outputs a sentolexicons object, which the compute_sentiment(). function specifically requires as an input; a sentolexicons object also includes a "[" class-preserving extractor function.
• The attributions() function outputs an attributions object; the plot_attribtutions() function is therefore replaced by the plot() generic.
• Defunct the perform_MCS() function, but the output of the get_loss_data() function can easily be used as an input to the MCSprocedure() function from the MCS package (discarded from Imports).
• Moved the parallel and doParallel packages to Suggests, as only needed (if enacted) in the sento_model() function.
• Sligthly modified appearance of plotting functions, to drop ggthemes from Imports.

sentometrics 0.4.0

• New functions: measures_delete(), nmeasures(), nobs(), and to_sentocorpus().
• Renamed following functions: any xyz_measures() to measures_xyz(), extract_peakdocs() to peakdocs().
• Dropped do.normalizeAlm argument in the ctr_agg() function (but kept in the almons() function).
• Inverted order of rows in output of the almons() function to be consistent with Ardia et al. (IJF, 2019) paper.
• Renamed lexicons to list_lexicons, and valence to list_valence_shifters.
• The stats element of a sentomeasures object is now also updated in measures_fill().
• Changed "_eng" to "_en"' in list_lexicons and list_valence_shifters objects, to be in accordance with two-letter ISO language naming.
• Changed "valence_language" naming to "language" in list_valence_shifters object.
• The compute_sentiment() function now also accepts a quanteda corpus object and a character vector.
• The add_features() function now also accepts a quanteda corpus object.
• Added an nCore argument to the compute_sentiment(), ctr_agg(), and ctr_model() functions to allow for (more straightforward) parallelized computations, and omitted the do.parallel argument in the ctr_model() function.
• Added a do.difference argument to the ctr_model() function and expanded the use of the already existing oos argument.
• Brought ggplot2 and foreach to Imports.

sentometrics 0.3.5

• Faster to_global().
• Set tolower = FALSE of quanteda::dfm() constructor in compute_sentiment().
• Changed intercept argument in ctr_model() to do.intercept for consistency.
• Proper checks on values of feature columns in sento_corpus() and add_features().

sentometrics 0.3.0

• New functions: diff(), extract_peakdocs(), and subset_measures().
• Modified R Depends from 3.4.2 to 3.3.0, and omitted import of sentimentr.
• Word count per document now determined based on a separate tokenization.
• Improved valence shifters search (modified incluce_valence() helper function).
• New option added for within-document aggregation ("proportionalPol").
• Now correct pass-through of dfm argument in ctr_agg().
• Simplified select_measures(), but toSelect argument expanded.
• Calculation in to_global() changed (see vignette).
• Improved add_features(): regex and non-binary (between 0 and 1) allowed.
• All texts and lexicons now automatically to lowercase for sentiment calculation.
• (Re)translation of built-in lexicons and valence word lists.
• Small documentation clarifications and fixes.
• New vignette and run_vignette.R script.
• Shortened project page (no code example anymore).

sentometrics 0.2.0

• First public release.

sentometrics 0.1.0

• Google Summer of Code 2017 "release" (unstable).