We should be using pkg_resources (or importlib.resources if our min Python version is 3.7) instead of uses of __file__.

$ get grep '__file__' sklearn/
sklearn/__check_build/__init__.py:    local_dir = os.path.split(__file__)[0]
sklearn/datasets/_base.py:    module_path = dirname(__file__)
sklearn/datasets/_base.py:    module_path = dirname(__file__)
sklearn/datasets/_base.py:    

Screenshot

Description

chart 3 dot menu is behind the chart title panel in chart maximize mode

Trying out a simple example using TuneSearchCV with LGBMClassifier and it fails on start.

Environment:

Python 3.8.3
tune-sklearn 0.3.0
ray 1.3.0
macos mojave 10.14.6

Code:

from ray.tune.sklearn import TuneSearchCV

from lightgbm import LGBMClassifier

lgmb_param_dists = dict(

                boosting_type=['gbdt','dart','rf'],
                num_leaves=(10,500),
 

Summary

The grayish background oval indicating a selected st.radio label has too much padding on the right hand side by a few pixels. Here's an example:

(Notice how the background rounded rectangle extends further to the right past "Notion" than it does to the left of the sel

The docs for IPython.core.interactiveshell.InteractiveShell.set_custom_exc have horribly mangled a warning message into a list of arguments. I can't work out at a glance why this is happening; it might be a sphinx.ext.napoleon bug, or a sphi

In recent versions (can't say from exactly when), there seems to be an off-by-one error in dcc.DatePickerRange. I set max_date_allowed = datetime.today().date(), but in the calendar, yesterday is the maximum date allowed. I see it in my apps, and it is also present in the first example on the DatePickerRange documentation page.

E

🚀 Feature

Log non-matching keys when loading checkpoints in non-strict mode.

Motivation

When load from an older checkpoint or partially initialize the model with pre-trained weights, we call the load_from_checkpoint API with strict=False, but we also want to know which keys are missing and unexpected.

Pitch

When [loading model states](https://github.com/PyTorchLightning/

When plotting plt.plot(np.ones(10), np.ones((10,0)) it raises a ZeroDivisionError, which confused me much.

Code for reproduction

import matplotlib.pyplot as plt
import numpy as np

plt.plot(np.ones(10), np.ones((10,0)))

This raises the error:

ZeroDivisionError: integer division or modulo by zero

Expected outcome

I think however, it should either r

In gensim/models/fasttext.py:

    model = FastText(
        vector_size=m.dim,
        vector_size=m.dim,
        window=m.ws,
        window=m.ws,
        epochs=m.epoch,
        epochs=m.epoch,
        negative=m.neg,
        negative=m.neg,
        # FIXME: these next 2 lines read in unsupported FB FT modes (loss=3 softmax or loss=4 onevsall,
        # or model=3 supervi

Is your feature request related to a problem? Please describe.
I typically used compressed datasets (e.g. gzipped) to save disk space. This works fine with AllenNLP during training because I can write my dataset reader to load the compressed data. However, the predict command opens the file and reads lines for the Predictor. This fails when it tries to load data from my compressed files.