We currently support reading FastText models from Facebook's format. The gensim.models._fasttext_bin does this.

This enables people to use gensim with a model that was trained using Facebook's binaries.

Sometimes, people want things to work the other way: they start with gensim, train a model, and then

One unit test in the R package is currently broken. Steps to reproduce on Mac

export CXX=/usr/local/bin/g++-8 CC=/usr/local/bin/gcc-8
Rscript build_r.R
cd R-package/tests
Rscript testthat.R

This results in the following error at the ends of the logs

[LightGBM] [Info] Saving data to binary file /var/folders/xq/wktq4zdx4jd3qdpk34d28m940000gn/T//RtmpiY1DzV/lgb.Dataset_1555

i'm a newbie in programming. I try to use this library. it's very useful for me.
I want to show centroid in K-means clustering. how to show it? thank u so much..

Would be great to have new option in Pool. Just like cat_features list of numbers or column names.

Describe the bug
When using the cdp driver, during closing of a browser page, this error sometimes appears.

{"level":"warn","time":"x","url":"x","error":"rpcc: the connection is closing","time":"x","message":"failed to close browser page"}
{"level":"error","time":"x","error":": rpcc: the connection is closing: session: detach timed out for session 5C391DF4E758E985AE3CBAA03774E562","t

We're undergoing an internal software audit and identified at least one textract component released under the Affero GPL: the EbookLib.

Lawyers are getting a bit antsy over this. In general, compatibility with GPL means that code released under a different license (e.g. MIT) and combined with GPL'd code must be released under GPL. This might create a b

Currently, we have a "drop_last_proba" parameter, which drops the last "probability" column in the feature set if it is set to True, because it is redundant: p(y_c) = 1 - p(y_1) + p(y_2) + ... + p(y_{c-1}). This can be useful for meta-classifiers that are sensitive to perfectly collinear features.

As mentioned by @bmreiniger in #527 , it might be useful to be able to choose whether the fi

Hello,

first and foremost, thank you for building this wrapper it is of great use for me and many others.

I have question regarding the evaluation:
Most outlier detection evaluation settings work by setting the ranking number n equal the number of outliers (aka contamination) and so did I in my experiments.

My thought concerning the ROC and AUC score was:

1. Don't we have to to rank th

The problem
Select Random Features preprocessor as implemented in the Preprocess widget allows the user to enter only a very limited number of attributes. I believe 99 is the current limit. The number of features on the input is usually larger. Say, in image embedding, we have thousands of features. The limit of 99 is therefore too stringent.

Solution
Remove any limit on the number of

When I click the Server Management button I receive an Internal Server Error
Error i noticed

I have been using the library for some time to parse my company invoices. I encountered that for my invoices I have line items which can be either of the two format. One way is that I create two templates file for each of it or if there is support for the multiple regex for lines and parser just picks the one for which match has been found.

Introduction
In case of complex and big data, it is not desirable to search using step=1, for the initial rough search it is much better to be able to search using a bigger step.

Description
The additional parameter should be introduced to the algorithm - step.

• Additional parameter for the Python implementation.
• Additional parameter for the C/C++ implementation.
• U