Update the citation in The R Journal.
The data.table has an update that affects the package, fix to make it work again. Details: akc ===== Error: package or namespace load failed for ‘akc’: object ‘update.dev.pkg’ is not exported by 'namespace:data.table'
Yes we have changed the name to update_dev_pkg(). This function upgrades the user's installation to the very latest commit of data.table in development. As such I'm surprised your package depends on it. It is supposed to be called by users, yes. But it is not supposed to be called by packages.
1.Using the latest version of roxygen2 to process the package submission, so as to deal with the CRAN notification. 2.Remove outdated URL in the document.
make_dict
function, ensure the words in lower case.lemmatize_dict
in the keyword_clean
function, users can now use a customized dictionary in this function.lemmatize_dict
and stem_lang
parameter for keyword_merge
function, so as to utilize other language and dictionary options when using 'textstem' package.keyword_vis
and keyword_network
function, the former code guides(fill = FALSE)
has changed to guides(fill = "none")
tidyfst::pairwise_count_dt
instead of widyr::pairwise_count
, so as to suppress the warning and run faster.keyword_cloud
to get single group of word cloud.doc_group
function to classify documents automatically.widyr::pairwise_count
,see https://stackoverflow.com/questions/63977452/warning-message-pairwise-count-function (keyword_group
).group_by
in the new dplyr version(keyword_vis
& keyword_network
).keyword_group
are removed because these links might be unstable. However, through the reference users could still find the source paper via searching.keyword_extract
.keyword_cloud
function for visualization of keywords.keyword_network
function to provide more flexible network visualization.keyword_extract
,put "id" first.keyword_merge
, iterate this process until no more "partof" method to make new changes to the data.frame.keyword_vis
to control transparency.Add a new option for keyword_merge
,namely "partof". This option could help merge tokens with relatively same meaning in the same document. For instance, "time series" and "time series analysis" would be merged into "time series". This method could only be applied to more-than-one-word phrases, which means "time series" and "time" would never be merged into "time" (which is considered as information loss). This alternative is an advanced method and should be used with caution.
keyword_extract
,give the alternative option to let parameter dict = NULL
, then the keywords would not be filtered by the dictionary. This could be useful in some circumstances.data.table
. This might be urgent. Details are listed in https://github.com/hope-data-science/akc/issues/1#issuecomment-572388600.keyword_vis
,change mutate(Group = str_c("Group ",group))
to mutate(Group = str_c("Group ",group) %>% reorder(group))
. This means that when there are more than 10 plots, it would be ordered as 1~10, but not 1,10,2...Fix the Description, adding quotes to API name. In the future the frame work will be extended to use more methods for keyword clustering in the tidy framework.
1.There might be two max values in keyword_merge, previously keep both, now randomly save only one.
2.Correct some typos in the vignette.
3.Document standardization,e.g. use TRUE and FALSE instead of T and F.
4.Update the description of package for more details.
5.Use \donttest{}
and \dontrun{}
for codes demanding longer time.This package could deal with relatively big data, but it demands computer with larger RAM.
Change back to unnest_token for high performance
1.Stop word should not be used in keyword_extract, hense I remove it. Because pre-defined stop words might hurt the following task of keyword extraction. 2.The previous keyword_extract did not implement the tokenizatin correctly (did not segment the sentence first, then extract n-grams), so I have adjusted the codes to make it right (there might be something wrong in the unnest_tokens in "ngram" mode, I use tokenizers::tokennize_ngrams directly to debug). 3.Add examples to make_dict function.
Export the tidygraph community detection functions. Change the detection function parameter name to "com_detect_fun". Add github url information and maintainer information.
0 errors | 0 warnings | 1 note
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.