gesistsa.difficulty is
deprecated. Instead, please use the respective frewweight
(for STM) and lambda (for Warp LDA). For legacy code
explicitly using the difficulty parameter, this change does
not break your code. However, if you use the following pattern, you
might need to change your legacy code accordingly.## This code is okay
set.seed(123)
wsi(abstracts_stm, use_frex_words = TRUE, difficulty = 0.8)
set.seed(123)
## You will get different results with oolong 0.5.0
wsi(abstracts_stm, use_frex_words = TRUE)
## You need to explicitly use the old default, which is quite high
set.seed(123)
wsi(abstracts_stm, use_frex_words = TRUE, frexweight = 1)The package-level documentation is removed.
Increase the required R version to 4.0 because of
keyATM
shinytest2miniUI as a dependency.export_oolong and deploy_oolong for
online deployment [thanks Marius Sältzer, Daniel Braby (and his friend
Louis), Johannes Gruber and Felicia Loecherbach for testing this
feature; thanks SAGE Ocean for the concept grant to support the
development of this feature]seededlda [thanks Marius
Sältzer]quanteda.textmodels
[thanks Marius Sältzer]wi, ti,
witi, wsi, and gsuserid as an suggested parameterupdate_oolong for updating object created by older
versions of oolongcli