I've noticed a declining trend in Excel usage among data scientists, especially those using Python/R, lately. What are your thoughts on this? Or is it still the most common tool? I really struggle to empathize with others on this. Constant errors, data loss, repetitive tasks—it's clear how shaky Excel's foundation is, but breaking old habits isn't easy. What do you all pay attention to when choosing tools?
Are data scientists becoming less dependent on Excel?
👁️ 4 views💬 2 replies❤️ 0 likes
2 Replies
I know Excel can drive you crazy with manual data fixes, forgotten saves, or one-click formula errors—trust me, I’ve been there. But when I switched to Python + Jupyter Notebook, for the first time, I could see *which line of code* was wrong instead of which cell. What used to take me two hours in Excel—building pivot tables—I can now rewrite in a few lines of Python and update in seconds. Plus, no more stressing over "did I forget to save the .csv?"—auto-save, debugging tools, the works. That said, I don’t think Excel will ever fully disappear: for quick client presentations, simple tables needing manual tweaks, or analyses that teammates can grasp instantly, Excel’s simplicity is unbeatable. Bottom line? Heavy lifting goes to Python, quick and visual tasks stay in Excel—best of both worlds.
Fully agree, mate. I've been using Excel for quick tasks for years, but it's getting more and more frustrating. The other day I tried cleaning a dataset with nested formulas across 500 rows, and when I exported it to CSV, the file got corrupted. Luckily, I had already moved most of the work to Python with pandas, but then I realized how much time I was wasting in Excel that could have been spent on what really matters.
The turning point for me was when a colleague showed me Jupyter Notebooks with reproducible code. Now I only use Excel for super simple tables or final reports, but for real analysis, I feel much safer with tools where I can version the code and repeat processes without fear of manual errors. That said, I’ll admit that in mixed teams (management + technical), it’s still the wildcard for “getting on the same page,” but even there, I notice the younger folks are demanding alternatives.