Python is not a great language for data science

The three most prevalent themes in this Hacker News discussion regarding Python and data science are:

1. Python's Dominance is Due to Ecosystem/Practicality, Not Inherent Superiority: Many users argue that Python's widespread use in data science is a result of its versatility, historical momentum, and strong library integration (the "batteries") rather than being the best language for the core data tasks themselves, especially when compared to specialized languages like R.

   • "Python doesn't need to be the best at any one thing; it just has to be serviceable for a lot of things." - lenerdenator
   • "The real reason python ended up with such good library support is they never really had a choice." - passivegains
   • "It's used in data science because it's used in data science." - dmurray

2. The Tension Between Exploratory Analysis (R's Strength) and Production/General Programming (Python's Strength): There is a clear divide noted between languages optimized for interactive statistical exploration (like R) and languages better suited for integration into broader software workflows or production systems.

   • "Productionizing R models is quite painful. The normal way is to just rewrite it not in R." - mohaine
   • "If you care more about logistics [file juggling, parsing, maintenance], your conclusion is pushed towards Python - which still does okay in the dataframes department." - jakobnissen
   • "If I want to wrangle, explore, or visualise data I’ll always reach for R. If I want to build ML/DL models or work with LLM’s I will usually reach for Python." - keeeba

3. The Lack of First-Class Tabular Data Structures in Mainstream Languages: A significant portion of the discussion centered on the oddity that mainstream general-purpose languages lack native, first-class support for the fundamental data structure of data science—the table or DataFrame—leading to reliance on complex, external library APIs (like Pandas or Tidyverse).

   • "Why aren't tables first class citizens in programming languages?" - RobinL
   • "The root cause seems to be that we still haven't figured out the best language to use to manipulate tabular data yet (i.e. the way of expressing this)." - RobinL
   • "This is like one of those people posting Dijkstra’s letter advocating for 0-based indexing without ever having read or understood what they posted." - kelipso (Referencing the fundamental differences in approach between general-purpose and data-focused languages.)