cacheAs now consistently covers all query
functions – frequencyQuery(),
corpusStats(), collocationScoreQuery() and
textMetadata(). The cache file records its relevant
production context: parameters, KorAP instance, index revision and
package version. cacheAsInfo() prints that record,
blessCacheAs() vouches for a file that is sound regardless,
and withCachedResults() takes the files as they are for one
expression, mode = "offline" refusing to compute anything
that is not in one already
the names of a named vc vector are now
consistently used as labels by frequencyQuery(),
corpusStats() and collocationScoreQuery(), as
collocationAnalysis() already did. Where a vector carries
no names, nothing changes: no label column appears that was
not there before
collocationAnalysis() now discards
collocates that occur less often than expected by chance.
logDice, by which it ranks and thresholds, expresses how
salient a pair is rather than how surprising, so a frequent word could
appear among the top collocates although the node does not attract it at
all: for Grund in a 5+5 window, Berlin reaches a
logDice of 3.84, close to triftiger at 3.96, while co-occurring
1.74 bits less often than expected. The new
minObservedExpectedRatio parameter defaults to 1 and keeps
such pairs out. Raise it to demand a stronger contrast, e.g. 2 for
collocates occurring at least twice as often as expected, or set it to 0
for the unfiltered result of earlier versions, e.g. to study repulsion.
collocationScoreQuery() is unaffected, as there the pairs
to score are given explicitly
changed logDice values:
logDice() is now computed as defined by Rychlý (2008),
14 + log2(2 * O / (O1 + O2)), so that its values are
comparable to those of Sketch Engine and other tools. It previously
multiplied the node frequency by the window size,
14 + log2(2 * O / (w * O1 + O2)), which added a count of
window positions to a count of word tokens and made the coefficient
asymmetric, so that swapping node and collocate changed the score.
Because w * O1 dominated the denominator for a frequent
node, rare collocates were penalised: for triftiger as a
collocate of Grund in a 5+5 window, logDice was 0.64, below
Berlin at 2.03, although Berlin co-occurs with
Grund less often than chance predicts. The values are now 3.96
and 3.84. Scores rise by up to log2(w), that is by up to
3.32 for the default context of 5 left and 5 right, so
collocationAnalysis() with the default
thresholdScore = "logDice" and threshold = 2
is now somewhat more permissive when recursing. Results computed with a
total window size of 1, as in the light verb construction example of the
Readme, are unaffected. The other association scores are unchanged: they
take the window size into account through the expected frequency
E, which is correct
changed ll() values: the
contingency table of ll() scaled only its row total by the
window size, an inconsistency spotted by Tim Feldmüller, leaving cells
that do not add up to one sample. Following Evert (2004), the sample
consists of co-occurrence tokens, so the sample size and both marginals
scale with the window: an occurrence of either word takes part in
window_size pairs. The expected co-occurrence frequency is
unchanged at window_size * O1 * O2 / N, which is why
pmi, mi2 and mi3 are unaffected,
but log-likelihood values differ, by 0.1% to well over 100% depending on
the frequencies and the window. This also removes the case where
N - window_size * O1 turned negative and the score became
NaN, since window_size * (N - O1) cannot: the
warning added earlier in this development version is therefore gone
again, having treated a symptom of this
improved coverage of the doc-prompting tests,
which now range from a frequency query over time to comparing collocates
across virtual corpora, keeping a result in a cacheAs file
and labelling corpora by name. They prompt current LLMs with the Readme
and check the code written from it, so that a gap in the documentation
shows up as a failing test. This guards the quality of the Readme and
improves vibe coding results. The approach is briefly described in Kupietz et
al. (2026)
fixed collocation analysis dropping snippets whose markup did not
have one particular shape, which cost about 15% of the hits of a
contains(<base/s=s>, ...) query: those are cut at the
sentence boundary and carry a <span class="cutted">
inside the match, and a match filling the whole sentence leaves an empty
context span. The two context spans are now read one by one and stripped
of their markup, whatever it contains (#14). This
only concerns servers that do not deliver tokenized matches, where
collocation analysis falls back to parsing the KWIC markup
findExample() no longer aborts with “replacement has
length zero” when a failed request leaves the query without any snippet
to take an example from (#14)
snippet2FreqTable() and
matches2FreqTable() work with their own default of an empty
stopword list again, which used to leave the table without the column
the stopwords are joined on
dropped the PTXQC dependency, which was imported for
two small string functions (lcpCount() and
lcsCount(), used by queryStringToLabel()) but
pulled in rmzqc, jsonvalidate and
V8, and with them the only dependency requiring a
libv8 installation.
cacheAs parameter to
collocationAnalysis() for transparent result caching: if
the specified RDS file exists, the cached result is returned
immediately; otherwise the analysis runs and the result is saved to the
file (.rds extension is added automatically if
omitted)max_delta_<score>,
winner<score>, loser_score<score>
columns etc.), including explicit winner/loser
webUIRequestUrl columns for association scores, ranks, and
percentile ranks. Missing per-label concordance URLs are now derived by
replacing the cq parameter of an available row URL with the
target label’s vc, and unsuffixed consensus
winner_webUIRequestUrl / loser_webUIRequestUrl
columns are populated when score-based URL choices agree. The names and
semantics of these columns may still change without a deprecation
cycleimputed, n_imputed and
imputed_<label> columns, and reported in verbose
mode, so that presence/absence artifacts can be told apart from measured
contrasts (filter(!imputed)). See the new “Interpreting
multi-VC comparisons” section in ?collocationAnalysis for
how to read the comparison columnscollocationAnalysis() now warns when repeated
node/collocate/label rows are reduced to their first occurrence for the
comparison columns, instead of dropping them silentlycollocationScoreQuery() now accepts a vector of
collocates and queries every combination of collocate and virtual
corpusKorAPConnection() is now a regular constructor function
instead of the S4 class generator, so that its manual page documents how
it is actually called. new("KorAPConnection", ...) and all
existing calls keep working unchangedfetchAnnotations() morphology so MarMoT and other
foundries keep all features from nested spans (#30)xml2, retaining
multiple lemma/POS values and improving robustness of token
alignmentfetchAnnotations() method added to
KorAPQuery class, to fetch annotations for all collected
matchesKorAPConnection(verbose=) default via
env var KORAP_VERBOSE or R option
rkorap.verbose (explicit argument still wins)ci functionauth() function (see ?auth)httr to httr2for all API
requestsggplot2 3.6.0 (thanks
@teunbrand)corpusQuery
resultsmergeDuplicateCollocates
functioncitation("RKorAPClient")matchStart and matchEnd columns to
corpusQuery results, that contain the start and end
position of the match in the textmergeDuplicateCollocates to merge
collocation analysis results for different context positionsquery to collocation analysis
results\link{} targets missing
packagecorpusStats methodspan parameter in
collocation analysis functionstextMetadata method, if
available, to retrieve also custom metadata for a text based on its
siglecorpusQuery method, if
availabletextMetadata KorAPConnection method to retrieve
all metadata for a text based on its siglewebUiRequestUrl column also to corpusStats
results, so that also virtual corpus definitions can be linked to /
tested directly in the KorAP UImetadataOnly is set, also tokenized snippets are
now retrieved in corpus queries (stored in
collectedMatches$tokens)%>%
(pipe function)hc_add_onclick_korap_search() now also works with bar,
column and pie highchartscollocationAnalysis function:
maxRecurse - apply collocation analysis recursively
maxRecurse timesaddExamples - If TRUE, examples for
instances of collocations will be added in a column
example. This makes a difference in particular if ‘node’ is
given as a lemma query.thresholdScore - association score function to use for
computing the threshold that is applied for recursive collocation
analysis callsthreshold- minimum value of thresholdScore
function call to apply collocation analysis recursivelydocumentation migrated to roxygen2md
Some examples in the documentation are now wrapped with:
## Not run
...
## End(Not run)This was necessary to meet the time limits of automatic CRAN checks under all circumstances and has otherwise no significance,
The change of the minor version from 6 to 7 automatically invalidates your cached query results. However, since the API has not changed, you can continue using the old cached results by copying or renaming your cache folder.
On linux, for example, this can look like this:
mv ~/.cache/R/R.cache/RKorAPClient_0.6 ~/.cache/R/R.cache/RKorAPClient_0.7or like this:
mv ~/.Rcache/RKorAPClient_0.6 ~/.Rcache/RKorAPClient_0.7To find the RKorAPClient cache directory for your environment, you can use:
R.cache::getCacheRootPath()collocationAnalysis method
(client-sided)randomizePageOrder to fetch result pages
in randomized orderignoreCollocateCase in
collocationScoreQuerywithinSpan (default:
base/s=s) in collocationScoreQueryverbose=TRUEsmooth added to
hc_freq_by_year_ciFixed some links in Readme.md
Removed redundant curl import
Fixed handling of single query term queries in highcharter-helper that had an empty plot and a warning
Unknown or uninitialised column: 'condition'.as consequences.
hc_freq_by_year_ci.demo("writtenVsSpoken")