Today I’m open sourcing “Grid studio”, a web-based spreadsheet application with full integration of the Python programming language.
This is awesome, thanks for building it! Have you thought about turning it into a SaaS service?
Small spelling correction: “… is so appealing to novice data scientists because of its untuitive behaviour.” should probably be “… is so appealing to novice data scientists because of its intuitive behaviour.”.
Thank you for the compliment. I have thought about turning it into SaaS, I’m not sure if there’s enough demand for it. At the moment I would like to develop it through open-source collaboration with a community of adopters. Maybe in the future it makes sense to offer a SaaS platform around it.
Good catch! I’ve edited it, but it doesn’t appear to propagate to the live version. Opened a post about it with the editor.
Thanks a bunch!
I’m a bit surprised that you find the code base clean, I personally had some issues with it. But I didn’t want to wait too long with shipping .
I did consider multiple language backends, but I felt that it would become too much for me to manage. It would require a better interface between the Go spreadsheet part and the language of choice, plus a whole lot more testing on my end.
Now that it’s in the open source realm I wouldn’t mind supporting more languages if someone has a good use case and a pull request
No way! Just yesterday I found myself using python and Excel and R to do some of the calculations and plotting. This is great! I would love to help you out continuing this project and I will!
Rick, you are a genius!
I am not data sciencist, just a guy who like to play with data but is not a master of spreadsheets and python, still know some of these and love both. So, Grid Studio looks like specially made for me
Thanks maciek! I hope it ends up being for you!
Quick question – do you do any sandboxing of the Python scripts? Wondering if I use this in a shared environment (e.g., with my colleagues/employees) whether they can access the general OS, etc. via Python.
There are two options currently: run it on your local computer or use the SaaS option that spawns your personal virtual machine on the DigitalOcean cloud environment (https://dashboard.gridstudio.io). In both cases your workspace Python session is isolated to the Docker container environment.
Does that make sense?
Second, Python has evolved from IPython to Jupyter Notebooks to JupyterLab. It enjoys a lot of popularity and rightly so, it offers a very nice work environment for data scientists with a strong emphasis on explainable code through long form Notebooks. Although, it lacks any kind of spreadsheet functionality that, in my opinion, is so appealing to novice data scientists because of its intuitive behaviour.
QGrid is one way to work with DataFrames as spreadsheets in Jupyter Notebook and JupyterLab
re: Jupyter and 2d projections of data
QGrid looks really interesting. I’m going to try it out for sure.
How can I install additional packages; say NLTK
You can use the terminal to install packages to the Docker environment.
Have you tried using pip3 install?
I already have these packages in my directory; when I use import it cant find the module. So I was wondering if there was a way to specify the path
I’ve tried importing file based packages in the Python interpreter before in Grid Studio and it worked for me. Are you sure you’re importing with the right path? Please double check the Python interpreter working directory with the os module and check the filesystem with the terminal.
Easiest would still be to run pip3 install nltk -U I guess
Thanks for your quick reply. let me try and see
This app is perfect for what I’m currently working on. Thank you, you beautiful human!