pandas
Python library providing DataFrame structures and tools for data manipulation and analysis.
Valued for interactive, exploratory analysis of semi-structured data12 and as a core part of Python's data science stack.345 The most common complaint is an unintuitive, footgun-prone API,678 and many commenters now prefer Polars or DuckDB.9101112
Why people recommend it
Python data ecosystem · 3
Seen as an essential piece of the Python data science stack alongside numpy, scipy and matplotlib.
“numpy/scipy/matplotlib/pandas/plotly/polars/pyarrow/jupyter, which is essential to my work”
“the open source ecosystem is great and matlab would be my go-to tool if they would not exist”
Show 1 more
Exploratory research · 2
Described as fantastic for iterative, interactive research and exploratory timeseries work.
“Pandas is fantastic for doing iterative and interactive research on semi-structured data.”
“I think it is the plotting ecosystem that keeps me in the Python universe.”
Data loading and cleaning · 3
Used to pull data from databases and clean or parse datasets.
“fill my local duckdb with datasets from each source with pandas.read_sql”
“I can use pandas to clean a dataset, but each cleaning task is usually one line of code.”
Show 1 more
“parses data (using pandas, python etc, not LLM but an LLM coded it)”
Familiar API · 2
Some users are comfortable enough with its API that they stick with it over Polars.
“I'm just too used to the Pandas API to use it as a replacement”
“am comfortable with pandas”
Production reliability · 1
One team runs thousands of lines of pandas on big data with rare production bugs, using schema wrappers and a type checker.
“the code base I work on has thousands and THOUSANDS of lines of Pandas churning through big data, and I can't remember the last time it lead to a bug or error in production.”
Caveats and dissent
Unintuitive API · 3
Commenters find simple tasks painfully unintuitive and say the API encourages bad habits and footguns.
“Pandas makes that kind of simple thing painfully unintuitive.”
“I used to spend my days unwinding bad habits DS pick up from years of panda abuse.”
Show 1 more
“enhanced ergonomics compared to Pandas and SQL”
Better alternatives · 3
Several commenters say Polars and DuckDB offer better syntax and performance.
“it is much better than pandas”
“I try to get my team off it but it’s an uphill battle.”
Show 1 more
“I used to spend my days unwinding bad habits DS pick up from years of panda abuse.”
Entrenched inertia · 2
Its entrenchment makes moving teams to alternatives an uphill battle, which some blame for stagnation.
“I try to get my team off it but it’s an uphill battle.”
“That inertia is not a good thing, and it's partially why there's stagnation in data science.”
Messy code · 1
Pandas notebooks often end up as hard-to-follow sequential wrangling.
“it is such a horrible tools”
Verbosity · 1
One user says a single DuckDB SQL statement can replace 40+ lines of pandas or Polars.
“I can use pandas to clean a dataset, but each cleaning task is usually one line of code.”
Internal design · 1
Even its original author has noted inherent shortcomings, according to one commenter.
“Pandas is terrific, yet even its original author has noted inherent shortcomings [1]”
Often compared with
Who recommends it in their own words
Threads where it came up
- Pandas Should Go Extinct 4 mentions
- Polars 2.0 3 mentions
- Pandas Exercises for Data Analysis (Interactive) 2 mentions
- DuckDB 2.0.0 LTS “Cinnamon Teal” 1 mention
- Pre-Release of Polars 2.0 1 mention
- Guide to data tools landscape for developers 1 mention
Also written as Pandas. Data: pandas.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 20
“fill my local duckdb with datasets from each source with pandas.read_sql”
“Pandas makes that kind of simple thing painfully unintuitive.”
“I'm just too used to the Pandas API to use it as a replacement”
“I can use pandas to clean a dataset, but each cleaning task is usually one line of code.”
“I used to spend my days unwinding bad habits DS pick up from years of panda abuse.”
“That inertia is not a good thing, and it's partially why there's stagnation in data science.”
“it is much better than pandas”
“I try to get my team off it but it’s an uphill battle.”
“enhanced ergonomics compared to Pandas and SQL”
“it is such a horrible tools”
“the open source ecosystem is great and matlab would be my go-to tool if they would not exist”
“am comfortable with pandas”
“Pandas is fantastic for doing iterative and interactive research on semi-structured data.”
“Pandas is terrific, yet even its original author has noted inherent shortcomings [1]”
“parses data (using pandas, python etc, not LLM but an LLM coded it)”
“I needed Pandas or numpy”
“What's worse is that Python libraries like pandas do that too.”
“the code base I work on has thousands and THOUSANDS of lines of Pandas churning through big data, and I can't remember the last time it lead to a bug or error in production.”
“numpy/scipy/matplotlib/pandas/plotly/polars/pyarrow/jupyter, which is essential to my work”
“I think it is the plotting ecosystem that keeps me in the Python universe.”