5 ms·

Python is probably mostly data science if I had to guess. jupyter notebooks have made python excessively used for data science.

mikestew 8y ago
Python is probably mostly data science if I had to guess. You think? I'll admit that my view of the Python community is super narrow, but I was thinking Django, me writing a bunch of test scripts, backend stuff. I know data science is a "thing" with Python, but I just didn't imagine that it had become it's main use. I guess TIL there's a whole other world of how people use Python, and it might be most of them. :-)
- laingc 8y ago
  I run an ML team whose primary language is Python, and I’ve been doing scientific computation in Python (eg finite elements, numerical optimisation) in Python for a long time. My personal impression is that Python is used in more fields than almost any other language, but that every community is only peripherally aware of other communities. For example, Python is big in Earth Sciences (or so I am told). Almost nobody outside Earth Sciences knows that. Until maybe 5 years ago, I had no idea that people used Python for web development. It had just never occurred to me. GvR famously had to google what wsgi was. So I think your final sentence probably applies to most users of Python.
  - milesvp 8y ago
    Yup. Python has some really good GIS libraries. So field that needs to deal with map data will often find itself drawn to python. And since it's a pretty newb friendly language you find that people who wouldn't otherwise program write simple scripts to manipulate the data and maybe output it to a csv so they can import it to excel.
- kevin_thibedeau 8y ago
  Python was popular (though smaller) before NumPy and everything dependent on it came about.
  - dagw 8y ago
    The first version of the project later renamed NumPy was released in 1995, so it's been around almost as long as Python.
    - fiddlerwoaroof 8y ago
      Yeah, I started using python around 2.0 and stopped using it because of 3, but from the beginning of my time using it, there always was numeric or numpy
- marcosdumay 8y ago
  At this point, Python is probably the second most used language for system programming (after C), one of the top 4 in web development (no idea on their ordering), around the top for other network protocols (I'd guess 3rd), the top language for data science, on the top 5 of scientific computing, on the top 5 for embedded computing (what sounds completely crazy, but there are few languages there), on the top 4 for game creation (probably 4th), the top one for electronics CAD and among the top for other kinds of CAD, the top one for GIS... I can't think of other niches right now, but whatever one you think, it's probably on the top 5 there.
- anothergoogler 8y ago
  Please don't project your narrow viewpoint onto PL communities.
edanm 8y ago
That's absolutely false. Firstly, while the advent of Jupyter notebooks have certainly made using Python better for data science, why would you think that would cause people using Python for unrelated purposes to stop using it for those purposes? Practically, while data science / ML / etc work is a big part of Python, I'd wager a much larger part is still web development work (Django et all). And of course, just using it as a "scripting" language, or using it to create GUIs, etc.
- davrosthedalek 8y ago
  To be fair, it is my impression that even in data science, python is more a less a scripting language, and all the libraries which do the heavy lifting computing-wise are written in something else, because speed. Has this changed?
  - pathseeker 8y ago
    The underlying libraries may be in C, but the vast majority of data scientists never need to alter the C so it's fair to say most data science code is written in python. Python/Numpy/pandas has enabled data scientists to avoid writing C entirely.
    - chasontherobot 8y ago
      about as much FORTRAN as C I believe
  - PeterisP 8y ago
    Sure, but for every line of code in such a library there's a hundred (or hundred thousand) lines of code that use that library. A small bunch of people maintain some CUDA code that's underneath a Python wrapper, but great many people write code that uses it.
kuang_eleven 8y ago
For as much as this is getting downvoted, I've been to PyCon several years running, and it is correct. While the actual usage of Python is very broad, the PyCon convention itself skews very heavily towards data science applications.
eksemplar 8y ago
I don’t know about that, Django is arguable still the most productive web-framework. I mean, I like JS as much as everyone else and graphql is better with Apollo than it is with graphene, but I can make a full django app and deploy it before I’m even done looking over the changes to my JS packages that have occurred since I used them last. Hell, even if you want to use vue or react, I still find django more productive than node, Apollo and prisma. And that’s not mentioning flask. So I’d wager that python powers a fairly large part of the web, and with good reason.

mikestew 8y ago

Python is probably mostly data science if I had to guess. You think? I'll admit that my view of the Python community is super narrow, but I was thinking Django, me writing a bunch of test scripts, backend stuff. I know data science is a "thing" with Python, but I just didn't imagine that it had become it's main use. I guess TIL there's a whole other world of how people use Python, and it might be most of them. :-)

laingc 8y ago
I run an ML team whose primary language is Python, and I’ve been doing scientific computation in Python (eg finite elements, numerical optimisation) in Python for a long time. My personal impression is that Python is used in more fields than almost any other language, but that every community is only peripherally aware of other communities. For example, Python is big in Earth Sciences (or so I am told). Almost nobody outside Earth Sciences knows that. Until maybe 5 years ago, I had no idea that people used Python for web development. It had just never occurred to me. GvR famously had to google what wsgi was. So I think your final sentence probably applies to most users of Python.
- milesvp 8y ago
  Yup. Python has some really good GIS libraries. So field that needs to deal with map data will often find itself drawn to python. And since it's a pretty newb friendly language you find that people who wouldn't otherwise program write simple scripts to manipulate the data and maybe output it to a csv so they can import it to excel.
kevin_thibedeau 8y ago
Python was popular (though smaller) before NumPy and everything dependent on it came about.
- dagw 8y ago
  The first version of the project later renamed NumPy was released in 1995, so it's been around almost as long as Python.
  - fiddlerwoaroof 8y ago
    Yeah, I started using python around 2.0 and stopped using it because of 3, but from the beginning of my time using it, there always was numeric or numpy
marcosdumay 8y ago
At this point, Python is probably the second most used language for system programming (after C), one of the top 4 in web development (no idea on their ordering), around the top for other network protocols (I'd guess 3rd), the top language for data science, on the top 5 of scientific computing, on the top 5 for embedded computing (what sounds completely crazy, but there are few languages there), on the top 4 for game creation (probably 4th), the top one for electronics CAD and among the top for other kinds of CAD, the top one for GIS... I can't think of other niches right now, but whatever one you think, it's probably on the top 5 there.
anothergoogler 8y ago
Please don't project your narrow viewpoint onto PL communities.

edanm 8y ago

That's absolutely false. Firstly, while the advent of Jupyter notebooks have certainly made using Python better for data science, why would you think that would cause people using Python for unrelated purposes to stop using it for those purposes? Practically, while data science / ML / etc work is a big part of Python, I'd wager a much larger part is still web development work (Django et all). And of course, just using it as a "scripting" language, or using it to create GUIs, etc.

davrosthedalek 8y ago
To be fair, it is my impression that even in data science, python is more a less a scripting language, and all the libraries which do the heavy lifting computing-wise are written in something else, because speed. Has this changed?
- pathseeker 8y ago
  The underlying libraries may be in C, but the vast majority of data scientists never need to alter the C so it's fair to say most data science code is written in python. Python/Numpy/pandas has enabled data scientists to avoid writing C entirely.
  - chasontherobot 8y ago
    about as much FORTRAN as C I believe
- PeterisP 8y ago
  Sure, but for every line of code in such a library there's a hundred (or hundred thousand) lines of code that use that library. A small bunch of people maintain some CUDA code that's underneath a Python wrapper, but great many people write code that uses it.

kuang_eleven 8y ago

For as much as this is getting downvoted, I've been to PyCon several years running, and it is correct. While the actual usage of Python is very broad, the PyCon convention itself skews very heavily towards data science applications.

eksemplar 8y ago

I don’t know about that, Django is arguable still the most productive web-framework. I mean, I like JS as much as everyone else and graphql is better with Apollo than it is with graphene, but I can make a full django app and deploy it before I’m even done looking over the changes to my JS packages that have occurred since I used them last. Hell, even if you want to use vue or react, I still find django more productive than node, Apollo and prisma. And that’s not mentioning flask. So I’d wager that python powers a fairly large part of the web, and with good reason.