One of the great strengths of the Python programming language is its wealth of third-party libraries. What good third party libraries have emerged in the Python world over the past year? In this article, I’ll introduce you to the top 10 Python development libraries that emerged in 2015. The scope of comparison here refers to new third-party libraries developed or created in 2015.

1. Keras

Keras is a highly modular neural network library written in Python that can be run on TensorFlow or Theano frameworks. The developers of Keras designed it to support rapid experimentation. Using the Keras library, you can dramatically reduce the time between idea and implementation.

2. yapf

Yapf is a Python file code formatting tool, but with a different algorithm than other similar tools. It grew out of clang-format, developed by Daniel Jasper. Basically, the algorithm takes the code and rearranges the original code so that it follows the best form of the code specification, even if it doesn’t violate it. The idea is similar to the Gofmt tool in the Go language, ending the “holy war” over formatting. If a project’s code base, whenever modified, is optimized through YAPF, the code style can be uniform, and there is no need to debate the style during every code review.

The ultimate goal of YAPF is to produce the same code as a programmer who follows the code specification. Can help you reduce the drudgery of maintaining code.

3. tqdm

TQDM (pronounced taqadum, تقدّم) means progress in Arabic. TQDM can add a progress message to a long loop. You only need to encapsulate any iterator. TQDM (Iterator) is a fast and extensible progress bar library.

Review images

4. pyvim

Vim editor implemented in Python.

Review images

5. snake

Snake is used to replace Vim’s VimScript for Vim’s plug-in programming. With the power of Python, it makes the plug-in programming more powerful than a tiger.

6. reverse-geocoder

A reverse Geocoder implemented in Python with strong performance and can be used offline.

7. pyxley

Quick development of data panels (Dashboards) using Flask and React.js. Displaying a data panel on a web page is the most intuitive way to share data science findings with others. For R, Shiny simplifies the work of data scientists developing web pages, and Pyxley is the Python equivalent of Shiny. Pyxley not only eliminates the need to write HTML and CSS, you can also add your own JavaScript to customize it.

Review images

8. pupy

Pupy is a remote Administration Tool that is open source and supports multiple platforms. Pupy also comes with a built-in Python interpreter that loads Python packages from memory and accesses remote Python objects.

9. tomorrow

Tomorrow provides an amazing implementation of the decorator syntax for asynchronous code in Python 2.7. The following is an example:

import time
import requests
 
from tomorrow import threads
 
@threads(5)
def download(url):
    return requests.get(url)
 
if __name__ == "__main__":
    start = time.time()
    responses = [download(url) for url in urls]
    html = [response.text for response in responses]
    end = time.time()
    print "Time: %f seconds" % (end - start)Copy the code

10. ibis

Ibis is a new project from Cloudera Labs, currently in preview form. It addresses the problem of cube size, but does provide a stand-alone Python experience that integrates with existing Python data ecosystems (Pandas, SciKit-Learn, Numpy). Future integration with machine learning and advanced analytics is planned.