-
Not sure what's happening here. The one thing that strikes me as odd is that it looks in python 3.5.2 directories, and then somehow ends up looking into a python 2.6 directory, and then the exception …
-
The user should be reminded of the restrictions on the documents that are being searched for collocations. TODO:
1. Provide a summary on the left
2. More??
-
-
Since the last version of Quanteda 0.99, I have lost the multi-threading on server with a virtual environment on docker. ( 1 out of 24 thread).
The multi-threading is working on my laptop.
-
These lines have a typo
# generic json - needs a textField specifier
mytf2
-
- extract noun phrases as in word2vec and the wordVectors package
-
@shadoof I made this change which seemed to greatly improve the memory handling (I noticed that many 'zombie' readers had freeCount == 4), but you definitely need to check it as I'm not sure it really…
dhowe updated
7 years ago
-
See https://meertensinstituut.github.io/mtas/search_cql.html on how to add `` as a range within which a collocation must be found.
-
The general stats gives the total number of documents for the filter/condition part.
Use the mtas 'stats' component to get the number of hits for the requested collocation.
-
Is it possible to facilitate n-gram in the word cloud such as "good call", "bad data" from the corpus, for say 2 to 3 or even more words?