We recently merged https://github.com/NicolasHug/scikit-learn/blob/master/examples/release_highlights/plot_release_highlights_0_22_0.py to summarise the highlights of 0.22, to be released in early November.

More release highlights, briefly demonstrating the new thing in action, belong in that example doc:

• Permutation based feature importance
• HistGradientBoostingClassifier with m

The code block at this location:

https://dev.pandas.io/docs/user_guide/reshaping.html#pivoting-with-single-aggregations

is not rendered correctly, as we accidentally try to render it as a note. The .. note:: should be replaced with .. code-block::

Describe the bug
POST to /api/database will overwrite an existing database instead of erroring

To Reproduce
Create a database with a POST to /api/database.
Run the same call again.

Expected behavior
Expect Metabase to error and say, "Database with name 'foo' already exists"

Actual behavior
Metabase gives another 200 and will overwrite any previous database config with

When building 1.30 it looked like I needed to add gettext-devel to the requirements on CentOS 7. I was going to PR the readme but was not sure if the additional dependency is correct or not. I figured I would ask first just in case.

The spec file I used as reference is the following:
https://src.fedoraproject.org/rpms/goaccess/blob/master/f/goaccess.spec

It'd just be nice to have links open in a new tab instead of killing my CC sesh.
(I'm using it to generate complex links on the fly based on input, so this is quite bothersome.)

Describe the bug
If you enter a string into a cell such as "3" and then use "toNumber()" to convert it to a number, toNumber() first tries to parse as a Long, and then tries parsing as a Double. So "3" will be stored as a Long, while "3.0" will be stored as a Double

However, if you edit a cell by using the single cell "edit" option and enter a number, and change the 'type' to Number, OR on

Currently we just take a shortcut in start_params where if the error terms are individual ARs, rather than jointly as a VAR, we still estimate a VAR and then take the diagonal elements of each coefficient matrix to get the AR parameters. But these may not lead to stationary models.

Related to #3997

Create a banner on the old documentation website that says the following:

You are not reading the most recent version of this documentation. 
<button>Go to the <latest> version .<button>

On the new website, addd a banner that says the following:

If you are using an older version of Pachyderm, go to archive-docs.pachyderm.io.

Auto-reviewers: @NiharikaRay @matthewwardrop @earthmancash @danfrankj

Hi,

I'm testing a knowledge repo implementation for our team using a git backend. Until know I could manage the deploy and Oauth using github, but I have an annoying problem of insistent and multiple request of my git user and password (no restart involved). I'm using the flask engine (and sqlite database). After the d

SMOTE-NC was implemented since #154 but it's not mentioned in the README (however it is mentioned in the documentation, although it's not quite easy to find). It would be good to add it along with a short description, as there are lots of use cases for this kind of algorithm that can mix both continuous and nominal variables.

It currently doesn't

For historical reasons, st.text/markdown/json/alert all share the same proto (Text) and they're all handled by the same element Text.tsx). This makes no sense, though. We should split them out.

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

Hello,

plotnine is a wonderful package. I am trying to figure out how to incorporate plots into a pyqt5 application. For matplotlib, there are a number of examples, but with plotnine, I don't know where to start. Any guidance is greatly appreciated.

Thank you.

I'm not sure how feasible this may be, but the old logparser program from Microsoft supported queries like

Select * from test.csv

And it would read from test.csv, without needing any config files.

Use mvn release:prepare to build the docs and copy them to the /docs directory