zilch42

Hi, I'm trying to use tq_transmute_xy with a window size of 5. I'm getting a vector length 5 being passed to my function for x but a vector the length...

control table and figure seq_id from yaml header or chunk options

3

The issue I am experiencing is that officedown uses different field code SEQ ids for tables and figures than word so if I create a word document with officedown and...

Empty Input logs

5

Hi there, I've just tried integrating shinylogs into my (rather complex) app and the resulting logs have data in Session, Outputs, Errors, but the Inputs table is empty. (not even...

logger.warning() formatting issue in topic_model.save()

2

Hi Maarten, I'm getting the following error trying to save a topic model without an embedding model pointer. ```python from sklearn.datasets import fetch_20newsgroups from bertopic import BERTopic # Documents to...

Words not in CountVectorizer vocab despite being well above min_df threshold

13

Hi Maarten, I'm having an issue with some important words not appearing within CountVectorizer when using `min_df` even though they are well above the set threshold. My understanding of `min_df`...

Typo in Merge Multiple Models docs

1

Hi Maarten, Thanks for the latest update, I think the latest merge_models function will be really useful for me! I think there is a typo in the min_similarity section on...

Hi there, just wondering if this package is still being maintained? It has disappeared from CRAN but it is still useful to my work.

Request: Zeroshot option to assign unassigned documents to outliers rather than reclustering

3

Hi Maarten, I have a use case at the moment where I'm using zero shot topic modeling to assign documents to a list of known clusters. I'm not really interested...

Why does `visualize_topics` use `n_neighbors=2`

5

Hi Maarten, Just wondering why the UMAP module in `visualize_topics` uses `n_neighbors=2` rather than something like 15? https://github.com/MaartenGr/BERTopic/blob/424cefc68ede08ff9f1c7e56ee6103c16c1429c6/bertopic/plotting/_topics.py#L79 A value of 2 results in a lot of topics sitting directly...

How to create layers for DataMapPlot

2

Hi Maarten, I've been following the new datamapplot addition which looks awesome. One of the great things about the interactive version is the ability to use layers that change as...