Supported?: Python, Jupyter Notebook, Postgres, MySQL, Airflow

[Mity]

I'm an aspiring coder. I have learned / am still learning the technologies mentioned above. For Python, I primarily use pandas, requests, matplotlib, SQL Alchemy. Are all of these fully supported on Apple Silicon? I currently have a 2019 16" Intel MBP. That machine is quite nice since it has 64GB RAM but it tends to struggle with very large Notebooks and for some reason, Zoom and Teams.

I'm not a professional developer at all. I'm still taking CS classes before (hopefully) starting an MS in CS next year. I don't think I'll be doing anything advanced like ML but I do want to learn mobile app development.

I'm thinking about getting an M2 MBP but only if all the software above is fully supported. If you use the tools above, can you please confirm? And can you install directly or do you need workarounds / package managers like homebrew / miniforge?

I have seen the following: https://forums.macrumors.com/threads/universal-and-native-apple-silicon-apps.2267176/

 

[BorderingOn]

What’s the processor in your 2019? My first thought is save your money until you truly understand what you need.

By now, any popular packages are most likely built for Apple silicon. You can check what you are wanting to use, e.g., https://pypi.org/project/pandas/#files. You can also build from source, if needed.

Different packages will have different install options such as pip, conda, disk image, etc. I don’t consider package managers “workarounds” and you’ll eventually find that they do serve a purpose.

 

[Mity]

What’s the processor in your 2019? My first thought is save your money until you truly understand what you need.

By now, any popular packages are most likely built for Apple silicon. You can check what you are wanting to use, e.g., https://pypi.org/project/pandas/#files. You can also build from source, if needed.

Different packages will have different install options such as pip, conda, disk image, etc. I don’t consider package managers “workarounds” and you’ll eventually find that they do serve a purpose.

Would I have to check for every package? What happens if there's a dependency that hasn't been resolved yet? Am I just imagining this problem or has it happened to you (or anyone else)?

 

[BorderingOn]

I’ve only run into it with some Docker images. But my needs aren’t super complex. I have seen others spending more time on it, like this: https://towardsdatascience.com/the-...ide-for-apple-silicon-macs-m1-m2-e9ef304a2c06

If you‘re mostly concerned about Teams and Zoom performance, I don’t think a new MBP is going to do much for you. If you haven’t already made up your mind, maybe go to an Apple Store and try installing everything 😀

 

[Mity]

I’ve only run into it with some Docker images. But my needs aren’t super complex. I have seen others spending more time on it, like this: https://towardsdatascience.com/the-...ide-for-apple-silicon-macs-m1-m2-e9ef304a2c06

If you‘re mostly concerned about Teams and Zoom performance, I don’t think a new MBP is going to do much for you. If you haven’t already made up your mind, maybe go to an Apple Store and try installing everything 😀

I'm more worried about Jupyter. I keep all my notes inside Notebooks and the one for my Intermediate Python class has difficulty loading, despite having a top spec Intel MBP. To be fair, I also have lots of other apps open at the same time, Word, YouTube instructionals and Zoom for class meetings. Zoom seems to push it over the top but the problem seems to be with how I store too much data in my Notebooks.

 

[altaic]

I run jupyterlab without issue (other than the usual conda nonsense).

 

[Xiao_Xi]

problem seems to be with how I store too much data in my Notebooks.

You may be creating new dataframes each time you modify a column. You can use Visual Studio Code to check what variables you have.

You should also change the type of the columns of the dataframes so that they occupy less memory space.

 

[Mity]

You may be creating new dataframes each time you modify a column. You can use Visual Studio Code to check what variables you have.

You should also change the type of the columns of the dataframes so that they occupy less memory space.

Believe it or not, it's not Pandas that's causing this. I take notes and do all of my homework in one Notebook per class. When I open up my coding projects that have Pandas, I've never experienced an issue with my Core i9 and 64GB of RAM. But all the code and formatting of notebooks, seems to be too much. Maybe this can't be fixed by switching to M1 Max?

 

[Mity]

I run jupyterlab without issue (other than the usual conda nonsense).

I also dislike conda; I always have dependency issues. That's why I did a standalone install. Awesome that it works perfectly on Apple Silicon!

 

[Xiao_Xi]

I also dislike conda; I always have dependency issues.

You may have dependency conflicts when adding a new package in a conda environment with some libraries installed. To limit them, you should list all the libraries to install when creating the conda environment.

 

[Mity]

You may have dependency conflicts when adding a new package in a conda environment with some libraries installed. To limit them, you should list all the libraries to install when creating the conda environment.

I was trying to use an R package from conda so I didn't expect any issues with the environment. My understanding was that I didn't need to create a new environment but I'm not certain. To be honest, I don't like using different environments since I don't have production level work. I removed conda and installed R as a standalone and added the R kernel into Jupyter, everything worked fine.

 

[Xiao_Xi]

I removed conda and installed R as a standalone and added the R kernel into Jupyter, everything worked fine.

If you use R, you can also use RStudio. It is more convenient than using Jupyter and R.

 

[Mity]

If you use R, you can also use RStudio. It is more convenient than using Jupyter and R.

I did. I used the included anaconda version.