Can't install QuantEcon notebook

Sorry for a newbie question: I was trying to install the QuantEcon package, but there is always this error message popping up

saying “can’t download Project. Make sure the URL is accurate.” I already downloaded the file from GitHub Desktop and made sure it was in the same folder as my .ipynb file. I try to run it on REPL as well, and also uninstalled and installed Julia again, but still couldn’t figure it out. Did I do something wrong here?

Thanks @Patrick_Wu Let me take a look to see what is going on and I will get back to you.

Typically the easiest thing to do is to clone the whole repository of the notebooks (i.e. ) which then doesn’t even do any downloads. Give that a shot in the meantime.

Are you on OSX or linux?

Hmm… I can’t replicate. This might be an OS specific issue? Can you give us more details on your setup?

Hi Jesse, thanks for the help! I did try to clone the whole repo of the notebooks, but it still says “Warning: Can’t download Project. Make sure the URL is accurate.”
I’m on OSX; I have Julia-1.5 (also Julia 1.2), GitHub Desktop, Jupyter is installed from Anaconda. The notebook is cloned to the same folder as the working directory, so this shouldn’t be an issue.

Sorry @Patrick_Wu not sure there is much I can help with. I think this could be an OS setup issue since it seems to be having trouble with curl? Although maybe there is a regerssion in the InstantiateFromURL.jl on OSX? @arnavs might know.

My suggestion: (1) see if curl works on its own, and if so; (2) delete the .julia folder on your computer entirely and try again. Make sure to use the cloned repository, which shouldn’t even need to download the github projects URL at all.

curl does work on my OSX, so this shouldn’t be an issue? I’ll try to delete julia entirely and try again ( I did it once but I guess maybe there are some folders I didn’t clean up). Thanks!

Yeah. The key is that .julia folder inside of your user folder.

Otherwise, so sorry, I don’t have OSX and am not sure how to test it.

Just gave this a go on my MacBook and can’t reproduce… here’s my setup:

julia> versioninfo()
Julia Version 1.5.1
Commit 697e782ab8 (2020-08-25 20:08 UTC)
Platform Info:
  OS: macOS (x86_64-apple-darwin19.5.0)
  CPU: Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz
  LIBM: libopenlibm
  LLVM: libLLVM-9.0.1 (ORCJIT, skylake)

and the latest InstantiateFromURL.jl (0.5.0).

A few thoughts.

  • The Julia command we’re using to do the download is, joinpath(pwd(), "Project.toml")), if you want to test outside the package.

  • It might be a firewall/security issue… I know OSX has some strict security settings, and maybe this process isn’t allowed to download without user input (or “run as admin” type stuff.)

  • For what it’s worth, my curl is the anaconda3 curl. Maybe they behave differently than the default (but skeptical.)

arnavsood@mbp:~$ which curl

Thanks, and please let me know what you find. It could be we need to update the instructions for OSX, etc.

Nevermind, it looks like Julia ignores the conda curl anyway, and uses the system one. Here’s a bit of the source code for

function find_curl()
    if Sys.isapple() && Sys.isexecutable("/usr/bin/curl")

Thanks! I checked the version info and it outputs exactly the same setup as yours (except that here it didn’t give the “Environment: JULIA_CUDA_SILENT = true” line; not sure why is this). When I use the Julia command you mentioned, it goes like this:

Didn’t give an error message but also no other output.

Hi @Patrick_Wu I think this might be something best asked on the julia discourse now that we have isolated it. Or, if you are on it, you could ask on the julia slack channel.

One thing that is useful when asking questions is to put the output and code as text rather than as images. See on quoted code.

@arnavs thanks so much for checking this!


Hi Jesse, thanks so much for the help!! I’ll try and check it out. In future questions (if they come up) I’ll keep that in mind!