Closed jsimonastro closed 1 year ago
What pandas version are you using?
1.3.1
Did you try creating a fresh environment and installing there?
I think the problem might be with isochrones or numba. This is because the last module to be called before the spiral into chaos was isochrones when interpolating, and isochrones uses numba (if I remember correctly...) under the hood.
The error is most likely being caused because of some corrupted temp file or something from isochrones
I'm trying this now, but haven't done it before. I defined a new environment called ariadne_environment, activated that environment, and then ran "pip install astroARIADNE". But it says it's already installed, so I must be doing something wrong.
I updated both isochrones and numba before posting the issue, and that didn't seem to change the error sequence.
What are you using to manage environments?
If you're using conda I could help you out
Yes, I'm using conda.
so, usually the instructions are
conda create -n name python=version
I assume you did so already so to activate the environment you go conda activate name
to switch to that environment.
The terminal should display something like this:
(ariadne) jayvains@Joses-MacBook-Pro ...
since my environment for developing ariadne is called ariadne.
If that's done then you should be installing stuff to that environment. If even then pip says you've already installed something, then pip is probably not installing to the anaconda path, in which case you should check that.
You can test out the new environment by opening a python terminal and trying to import ariadne. It should fail
I've created and activated the environment, but pip install still says that ariadne is already installed, and I can import ariadne in python.
Then, as I said, pip is probably not installing packages in the environment but instead at default. You should check your PATH variable and everything related to that.
I can't help you with that.
I used "pip install --target /opt/anaconda3/envs/ariadne_environment/" to install ariadne and all of the dependencies to the new environment. However, when I start python and import ariadne, it still uses the original installation (/opt/anaconda3/lib/python3.8/site-packages/astroARIADNE) even though I'm in the new environment. If I remove /opt/anaconda3/bin so that the only anaconda reference in my path is /opt/anaconda3/envs/ariadne_environment/bin, the same thing happens. Do you have any advice for what should be in my path to direct python to the correct installation/environment?
I can share my PATH variable, maybe it could help you figure out what's going on. Though anaconda should automatically change stuff up
/Users/jayvains/.rvm/gems/ruby-3.0.0/bin:/Users/jayvains/.rvm/gems/ruby-3.0.0@global/bin:/Users/jayvains/.rvm/rubies/ruby-3.0.0/bin:/Users/jayvains/bin:/usr/local/bin:/Users/jayvains/opt/anaconda3/envs/ariadne/bin:/Users/jayvains/opt/anaconda3/condabin:/usr/local/bin:/System/Cryptexes/App/usr/bin:/usr/bin:/bin:/usr/sbin:/sbin:/Users/jayvains/.rvm/bin
In that example you can see /Users/jayvains/opt/anaconda3/envs/ariadne/bin
because I'm already in that environment, so something like that should pop up.
Also executing which python
in the terminal should output something like this: /Users/jayvains/opt/anaconda3/envs/ariadne/bin/python
Ok, I've got it working now. The problem was that I created the new environment with nothing in it. I then tried using pip to install everything needed for ariadne, but since that environment didn't have pip in it, it was using the pip from my main conda, which wasn't putting things in the right place. So instead, I used conda install to put some of the basics (python, pip, etc.) into the new environment. Then I was able to use the new pip to install ariadne and its remaining dependencies into /opt/anaconda3/envs/ariadne_environment/lib/python3.1/site-packages/ without having to adjust my path or any target directories by hand.
This sidesteps the original issue that I raised, but it's still not clear what the incompatibility was or how it arose!
If that solved the issue, then my first impression was correct: There were some corrupted temporal files. They could've been from ARIADNE, but the part that raised the error was isochrones during the interpolation.
You could try to look into the issue further by actually debugging the code, but that would be 100% up to you. As far as I'm concerned the issue is solved.
If you need anything else, don't hesitate to ask.
Hi @jvines, I used astroARIADNE (v1.0.4) successfully a few months ago, both following the example documentation at https://github.com/jvines/astroARIADNE#how-to-use and then repeating for another star I was interested in.
Now I am trying to run SED fits for some additional stars and am getting the following error output when I execute f.fit_bma():
I have tried updating to v1.0.9, updating all of the dependencies referenced above, and removing the .pyc files in pycache, but none of those have had any effect. I haven't been able to figure out what zip file is the source of the error. Can you help me understand where this error is originating? Thanks!