So basically I started learning python and with Visual Studio Code. It was great till I had to install simple packages(pandas, numpy, etc). I couldn’t get it to work because it was too complicated to change path and all that as I’m a beginner.
This is why I installed Anaconda, like all the modules were there. What I’m worried is about that in the future if I work in a company can I still use anaconda?
Is it common for programmers to use anaconda for module management?
Yes. Many enterprises use Anaconda. Anaconda continues to be a very relevant tool in managing python environments. Many tools such as AzureML and other build conda environments as a core component of their backend processes.
You can also try:
poetry
pip
Related
Im facing the following situation:
1. I wish to use TF 1.12 with MKL on windows
2. I have to use pip (our system supports only it so far)
problem is that currently Im unable to compile TF1.12 on windows (only on linux..).
I found that Conda has it right out of the box! so I was thinking
"hmm lets take it from Conda, probably someone has thought about it :)"
never the less, havent find any package or program that can do it, so I've decided to use Stack-overflows' power to look for the solution.
any help in this subject would be really appreciated
I want to download Julia (the last version is 1.0) from Anaconda. However, you can download from https://julialang.org/. My questions are: What are the differences between both ways of installing Julia? Can I install, for example, DifferentialEquations.jlor Symata.jl without problems if I choose Anaconda? If I choose Anaconda, how good is the package management?
The only benefits of Anaconda are, as far I'm aware, that it automatically selects the right binary (i.e., OS), and likely has a slightly easier updating experience (for the language itself). However, it does not seem to support Windows (https://anaconda.org/conda-forge/julia), so if you happen to be on that platform, you are out of luck. I would recommend grabbing the binary from the website directly, the installation process is very straightforward.
The management of Julia packages will still happen inside Julia. Julia 1.0 has a very good package manager called Pkg. You can read more on installing packages within Julia on https://docs.julialang.org/en/v1.0.0/stdlib/Pkg/.
Image of installed Anaconda Directories OK, long story short. I just started learning about CoreML for iOS. So from what i gathered, theres only about 4 trained models on Apple's website. In the search of some more robust models, i found Caffe was a hugely mentioned one. You need to use python to be able to translate some of these other 'external' models. I installed Anaconda via the continuum website for mac and am using the only supported 2.7 python language with keras/scikit/and a few other modules....... So from what I think it could be is after I installed anaconda at Applications level on my mac, it should of been originally installed at the User level. I reinstalled it at the User level, so technically I have two instances of anaconda on my system. Neither will work and inside Anaconda, i cannot import coremltools.
I have a hunch that maybe i need to change some pointers. Or just uninstall it completely from the machine and reinstall it at the User level from the start. I have about a year of swift experience, so im much more confident in my swift ability vs terminal ability. If anyone could coach me through how to cleanly remove it from my macbook pro, and explain the commands, I would greatly appreciate it. Or give any insight about whats going on or what im not doing. I can go into more detail if need be. Cheers in advance
I have Python 3 on OSX Yosemite installed with Brew. I want to make some simple charts (like in Excel). I run Python from bash. Which library should I investigate, and the proper way to have it installed? I am looking for library which is easy to install, and then easy to use. Output probably png files?
matplotlib is a good choice.
How to install it on OS-X is mentioned in the installation FAQ.
Especially in combination with IPython notebooks it gives you a nice interactive environment to play with it.
Another alternative would be gnuplot. Pre-built packages are available.
It is not tied to Python, it is basically a domain specific language for graphing. There are python bindings available, gut it is also not difficult to use it with subprocess.
If you can use commercial libraries, you may also consider ChartDirector.
I have noticed that often, MacPorts will install dependent software that is already embedded into my OS (OSX). Particularly, Python is a frequent example of this.
Is the MacPorts software so intelligent that it will check for native installations and install dependencies ONLY if the native installation is not compatible with the port I am trying to install?
Or, does it simply see in the list of dependencies that, "Aha, I need Python, and Python is not installed as a port, so I will install it" (even though there is a native installation of Python)?
Note: This question is meant generally, not only regarding Python. Just using that as an example.
The package systems usually distrust the pre-installed software and install their own. It’s easier to control it, you get a more recent versions of the software and also have greater control over versioning, which helps with dependency management.