Hey #Nostriches!

I'm an engineer currently studying data science and I've always used Windows. I'm thinking about switching to Ubuntu for better speed and because it seems more compatible with the frameworks we use in data science. I discover yesterday that VsCode works on Linux so I have no fears anymore. 😅

- Do you think it's worth it? 🤔

- How has your experience been using Linux for data science? 🧑‍💻

- What's make you be a Linux user instead Windows? 🐧

- Also, would I be able to work with Excel files without any issues using LibreOffice? 📚

I'd really appreciate some help from any Data Scientists, Data Analysts, or Machine Learning/Deep Learning professionals/Developers who use Linux.

Any feedback and personal experiences would be greatly appreciated! 🤙

Ps: I don't gaming or care about it.

#asknostr #datascience #machinelearning #deeplearning #linux

Reply to this note

Please Login to reply.

Discussion

I think learning the linux commadline has several advantages, since most servers are powered by Linux. For sure in many jobs they will be happy to hear you have Linux experience.

I would recommend trying Linux. However depending on the complexity of the excel files you are working on, libreoffice will struggle to be a good replacement. You might be able to get away with a windows VM in your Linux machine to run excel in or the web version of excel.

I've heard that opening .xls files in LibreOffice can sometimes result in truncated data, depending on how the spreadsheet was created. .csv files, however, seem to work fine.

I've done some research, and everyone seems to recommend using a virtual machine (VM) in this situation. Since we only need these files in the initial stage for importing into a Jupyter Notebook and performing exploratory data analysis (EDA), I think a VM is a reasonable solution and maybe never need to do that.

Thanks, mate 🤙

I used WinApps when I tried to solve this problem: https://github.com/Fmstrat/winapps

Unfortunately for me my machine couldn’t quite perform well enough running the apps in a virtual machine, so it wasn’t useable for me.

Totally worth it. You’ll learn so much and it will make you a better engineer overall!

https://youtu.be/_H4E6eTRHAo?si=EsKylj5S4qlTBura

the big advantage of linux is optional privacy and security. That‘s probably not relevant for your task. Maybe analyzing data in a VM without network access in Qubes OS - after you got comfortable with ordinary linux distributions.