Skip to content

Latest commit

 

History

History
39 lines (36 loc) · 2.33 KB

start-pro.md

File metadata and controls

39 lines (36 loc) · 2.33 KB
Raw

Getting started instructions for using JuliaPro and MIRT

  • Install JuliaPro 1.1 (or later) from https://juliacomputing.com/products/juliapro.html

    • Email registration is needed
    • Refer to the Quickstart Guide for prerequisites and for the REPL layout
    • When prompted about the Pkg server, leave blank to use the default.

  • Explore the extensive Julia documentation at https://docs.julialang.org/

  • Launch JuliaPro and type 1+2 at the Julia REPL prompt to verify it works.

  • Use the ] key at the Julia REPL to enter its package manager (pkg).

  • Type ? and press Enter at the Julia pkg prompt to peruse the many pkg commands.

  • Add needed packages using the add command.

    • Specifically: add MIRT Plots IJulia Debugger FFTW FFTViews Arpack
    • The package IJulia is needed for Jupyter notebooks.
    • FYI: add MIRT automatically loads from https://github.com/JeffFessler/MIRT.jl because MIRT.jl is a registered package.

  • Type precompile to have Julia precompile the added packages. (This will save time later.)

  • After you are done adding packages, press the backspace key to return to the REPL prompt.

  • Later if you need to add more packages just type the ] key again at the REPL prompt to enter the package manager.

    • Julia is under active development so code is updated frequently. It is wise to type up (short for update) at the package manager prompt every week or so to get automatic updates of any packages you have installed.
    • After running up, you can type gc (for garbage collect) to have Julia remove old versions of packages.

  • For some Julia tutorials see http://web.eecs.umich.edu/~fessler/course/551/julia/tutor/

  • For some signal processing demos in Julia see http://web.eecs.umich.edu/~fessler/course/551/julia/demo/

  • Test your MIRT installation by displaying a test image:

using MIRT: jim, ellipse_im
x = ellipse_im(100)
jim(x, title="test")

  • This test should produce a grayscale image of the famous Shepp-Logan phantom in the Plots tab of JuliaPro.

  • To learn about the jiffy image display function jim, type ?jim at the REPL.

  • JuliaPro uses Juno and Atom and there is lots of online docs about these. I use the vim-mode-plus key bindings, installed using Atom preferences.

  • To start a Jupyter notebook, type using IJulia; notebook()