Pandas feels clunky coming from R. What about Haskell?

mchav.github.io

26 points·mchav··7 comments

Although not quite the same without a pipe operator, https://pola.rs was an improvement for me when missing the R dataframe syntax.

OP makes that concession in the first section of the post. (I may or may not have made a similar comment before deleting in kneejerk shame)

I find pandas clunky as well, and I’m not a fan of SQL. So I started this DSL https://github.com/nealbob/pivotal-py

> You need to read the previous line to understand what

that ended rather abruptly?

> This has a great SQL-ish API. Python is similar but starts to be a little clunky since it requires you to think about indices:

groupby has an as_index parameter for this very purpose

> Deducting the discount

You focus on doing the subtraction during the group by. Is there any good reason for this? You could either do it as a step before, or after summing up both columns. Putting too many things into one command is not good practice yet you benchmark the language based on how easy it is to do said bad practice

I think the original author picked this example to broadly illustrate how easy it is to make ad hoc changes to your query without worrying about lot about implementation details. Polars, for example, converges on a similar API and gives you the flexibility. You can iterate then refactor easily later to what you consider good practice.

For me the whole piping felt like making everything less readable and harder to debug compared to a string of commands.

Reply on news.ycombinator.com