Report: Python 2 not maintained, however nonetheless used
Regardless of Python 2 nearing end-of-life on January 1, 2020, 10% of Python builders had been nonetheless utilizing it in 2019, in accordance with JetBrains’ 2019 Python developer survey.
The share of builders nonetheless utilizing Python 2 has been reducing year-over-year. JetBrains’ 2017 survey discovered that 25% had been utilizing Python 2 and their 2018 discovered that 16% had been nonetheless utilizing it. Whereas this report incorporates knowledge from 2019 and Python 2 reached finish of life in 2020, an ActiveState survey from earlier this 12 months revealed that 50% of corporations didn’t have a plan in place for Python 2’s finish of life.
The commonest use for Python 2 was internet growth, whereas the most typical use for Python three is knowledge evaluation. “Though Knowledge evaluation is extra well-liked amongst Python builders, it’s attention-grabbing to see that its share amongst those that use Python 2 is decrease than internet growth’s share. That is most likely as a result of knowledge evaluation in Python has grown extra well-liked lately, whereas internet growth is a extra mature area and a few internet builders have a number of legacy code to take care of,” JetBrains wrote within the report.
RELATED CONTENT: Shifting from Python 2 to Python three
The commonest use for Python usually isn’t shocking. Fifty-nine p.c of respondents stated they use Python for knowledge evaluation. This was adopted by 51% utilizing it for internet growth, 40% utilizing it for machine studying, and 39% utilizing it for DevOps, system administration, and writing automation scripts.
Python builders don’t are inclined to solely use Python for one objective. The respondents had been requested “What do you employ Python for?” and the imply variety of functions was three.9.
The report additionally regarded into the most well-liked Python frameworks and libraries. Flask and Django had been the most well-liked internet frameworks by far, with 48% and 44% of builders utilizing them. The subsequent closest was Twister at 5%. There was a extra even distribution of recognition for knowledge science frameworks and libraries. NumPy was the most well-liked, with 64% of Python builders utilizing it. This was adopted by Pandas (55%), Matplotlib (46%), SciPy (36%), SciKit-Be taught (33%), TensorFlow (26%), and Keras (20%). All different knowledge science libraries and frameworks had lower than 20% use.
JetBrains gathered info from 24,000 Python builders from greater than 150 international locations for the survey.