Skip to content

Latest commit

 

History

History
53 lines (28 loc) · 5.39 KB

Anaconda.md

File metadata and controls

53 lines (28 loc) · 5.39 KB

Anaconda Scientific Python installation

Python is available from several sources, and comes in two flavors: older Python 2.x and newer Python 3.x. The PAW pipeline uses Python 3.x (3.6 at the time this was written). Python is an expressive, modern computer language, and has become the top computer language. Python includes an extensive standard library that adds much functionality. What is included in the standard library is very conservative and many powerful packages for scientific applications are, unfortunately, not included. The challenge of maintaining several Python extension packages has resulted in scientific Python distributions, such as Anaconda. The Anaconda Python 3.6 distribution is recommended for the PAW pipeline.

The first pipeline steps of converting RAW files and running Comet require Windows 7 or Windows 10 systems, so Anaconda 3.6 for Windows (typically 64-bit) has to be installed on the computer where these processing steps will take place. If other computers are used to perform any downstream PAW pipeline steps, they will also need Anaconda 3.6 installations. There are issues with non-Windows systems accessing NTFS-formatted hard disks. Files can be often be read but not written. Files can be copied to non-NTFS formatted volumes, or drivers like Paragon NTFS for Mac can be used.


CRITICAL MESSAGE

Newer Anaconda distributions break the histo_GUI.py script

At the moment, an older Python 3.7 distribution of Anaconda still works. Changes were made to Matplotlib in recent versions that gave warnings when running the hisot_GUI.py script. Those have finally become errors in the most recent distributions.

You can get the older Python 3.7 distribution >>here<<. The file to download is Anaconda3-5.3.1-Windows-x86_64.exe at the bottom of the web page.


Downloading Anaconda

The Anaconda download link has options for Windows, Mac, and Linux versions, and for Python 3.6 or Python 2.7 versions (as of 2018-04-06). You should get the Python 3.6 download:

Anaconda_download


Launching Python 3

Follow the installation instructions for your platform. On Windows, Anaconda programs will be added to your start menu. There are many different ways to run Python programs on most systems. Python tutorials and books usually have some discussions of the different ways. This short video covers a few ways. Anaconda has online documentation about the distribution and the Anaconda Navigator GUI environment. The Spyder integrated development environment (IDE) is fine for launching scripts and great if you want to do some Python coding.

I prefer an IDE called IDLE that is part of the Python standard distribution. IDLE has been around for years and does the basics in a clean way. Since Anaconda includes more modern IDE tools, figuring out how to launch IDLE is not so obvious in Anaconda. It is best to launch IDLE from a command line environment with some additional Python paths defined, called Anaconda Prompt:

Anaconda_start_menu

In the console window that will launch, type IDLE and hit the RETURN key.

Ananda_console

The IDLE console window should appear:

Idle_console

The Python version information should indicate an Anaconda distribution and a Python 3.x version number. See the documentation on IDLE for a rather dry description of features. There are other tutorials, and also shorter and longer YouTube videos available. Keep in mind that IDLE is a more integral part of the Python distributions from python.org and is easier to launch with those distributions. IDLE opens with a single Python console window where you can execute Python statements. Existing Python programs can be opened from the File menu. The console window is where the standard input, output, and error messages go. The IDLE features that I like are that you can have multiple source code windows open. There is nice statement highlighting, indenting, and auto completion. The interactive debugger is also nice. An advantage over running from a Windows command shell is that console (program) output will persist.


Keeping things up-to-date

One of main the reasons to use a scientific Python distribution like Anaconda, is maintaining all of the installed packages. Anaconda includes the Conda package manager and Anaconda should be updated occasionally. This is best done with two command line calls, the first updates Conda itself, then Anaconda is updated:

Conda_update

Anaconda_update

If you get error messages about conda not being defined, make sure you are launching an Anaconda Prompt command window.