[ANN] DiffEq and friends now available on JuliaBox

juliabox

#1

A quick note to say that the DifferentialEquation packages are now installed by default on Juliabox. A huge shout out to Chris and his collaborators, for making Julia by far the best language to work with differential equations of all kinds.

As you know, this is a large set of packages, with many dependencies. If you have installed DiffEq packages manually, or even otherwise if you find instabilities with dead kernel messages etc, please re-install your packages. Go to the package management UI, add or delete packages if appropriate, and click Start.

As always, please let us know on juliabox@juliacomputing.com if you have any questions, kudos or brickbats.

Regards

Avik (on behalf of the Juliabox team)


#2

I may start calling them “friends” instead of “dependencies”. Sounds much more inviting.

Nice work!


#3

Hi @avik thanks this seems like a great choice.

May I report unexpected behavior here? After my last build on JuliaBox, these are the reported versions:

Julia Version 0.6.0
Commit 9036443 (2017-06-19 13:05 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  CPU: Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz
  WORD_SIZE: 64
  BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Haswell)
  LAPACK: libopenblas64_
  LIBM: libopenlibm
  LLVM: libLLVM-3.9.1 (ORCJIT, haswell)

62 required packages:
 - Atom                          0.0.0-             non-repo (unregistered)
 - AutoGrad                      0.0.0-             non-repo (unregistered)
 - AutoHashEquals                0.0.0-             non-repo (unregistered)
 - BusinessDays                  0.0.0-             non-repo (unregistered)
 - CSV                           0.0.0-             non-repo (unregistered)
 - Calculus                      0.0.0-             non-repo (unregistered)
 - Clustering                    0.0.0-             non-repo (unregistered)
 - Conda                         0.0.0-             non-repo (unregistered)
 - DSP                           0.0.0-             non-repo (unregistered)
 - DataFrames                    0.0.0-             non-repo (unregistered)
 - DataStreams                   0.0.0-             non-repo (unregistered)
 - DataStructures                0.0.0-             non-repo (unregistered)
 - DecisionTree                  0.0.0-             non-repo (unregistered)
 - DifferentialEquations         0.0.0-             non-repo (unregistered)
 - DistributedArrays             0.0.0-             non-repo (unregistered)
 - Distributions                 0.0.0-             non-repo (unregistered)

And goes on like this. JuliaBox notebook seems to work normally never the less.


#4

Yeah, we know that report is incorrect, working on fixing it.


#5

And it should be fixed now.


#6

A PSA that Juliabox has been upgraded to version 4.0.0 of DifferentialEquations. This is a major upgrade to the DiffEq family of packages, since it contains a series of backward incompatible changes. Please update your code if you’ve used DiffEq on JuliaBox.

As an aside, Dolo.jl is now available on Juliabox.

Once again, if you see any kernel crashes, that means your custom installed packages are now incompatible with the pre-installed packages, and you need to remove or re-install your custom packages. There is now a big Reset button in the package builder UI that makes it easy to remove any custom packages, and get your instance back to a clean, pre-installed setup.

Regards

Avik