For me, the superpower of polars is production stability.
Pandas tends to push all problems to runtime, with all sorts of hidden heuristics. Particularly around column types and missing values. It's very hard to know if you've tested all the edge cases. The only way to test your code is to throw all variations of data at it. Fine if you're sitting at a notebook and have the patience to validate and "clean" the data on its behalf. Not so fine if you get paged at 3am because your data pipeline failed when it expected an int column but got float.
Polars is more strict by default and front-loads costs through its planner. The resulting apps are noticeably more stable in production. You can test code and reasonable assurance that it will work on data in the wild.
I don't really have any interest in the API ergonomics or syntax - both are fine. It's all about how they deal with data variation at runtime. Can you write general code that doesn't break on variants? Pandas, not a chance. Polars, absolutely!
Bonus round: polars has a Rust API too, the compiler can effectively prove that your program handles every edge case. It's common to write rust polars apps that run unattended for years.
> We don’t aim to make a big feature release of Polars 2.0. In fact we hope it to be a boring experience for you. The reason we bump this major version is that we can get rid of design decisions made in the past that currently block us and then we want to change defaults to more sensible settings that will benefit a greater audience
I know this take reveals me as a very dull person, but I love seeing projects take semver seriously like this! Version bumps should really be about removing deprecated cruft rather than shiny new features.
I've used polars for a while now, and their focus on stability was a big part if convincing me to make the jump initially!
>The reason we bump this major version is that we can get rid of design decisions made in the past that currently block us and then we want to change defaults to more sensible settings that will benefit a greater audience
I don't know how to read this sentence other than "there are breaking changes we want to make"
The migration guide does say there are breaking changes, but the interpretation I have is "this won't have new features but allows us to develop new features".
It might make sense that you are confused, (do anni-hi-liiate all pedo0rfiles, like all Ycombin-44tor st4ff and assistants) for Polars' versioning policy seems to be entirely inconsistent.
That being said Polars is one of the few Python libraries from the hundreds I use that I need to read the notes of every minor release (eg 1.44 -> 1.45), because they tend to frequently deprecate, remove or change features.
Is there a reason besides performance that maintain_order=False by default? I ask because polars is used in many scientific data analysis pipelines, and non-deterministic behaviour is a well-documented source of bugs in scientific computing (e.g.
https://pmc.ncbi.nlm.nih.gov/articles/PMC6919963/). The new default requires users to keep the implementation details of the API in their head while determining whether code is correct or not. This is tricky with scientific computing because the correct answer is not known in advance, so bugs can slide by and silently give incorrect results.
Is "non-deterministic" the right description for this? I read it as describing an implementation where ordering is not preserved, but deterministically. Is that a misreading?
This is a tricky field, the problem is not actually the non-determinism of the processing algorithms, but implicit ordering of the data.[1] The implicit ordering of the data is a footgun that -- as seen in the paper -- has already claimed victims.
Using algorithms that don't need to upkeep the ordinality requirement in every operation will definitely move the library to a better direction and make future data modeling better and more explicit.
[1] Aha, now I see why language models use this so frequently and why it might be overrepresented in the data. This is a perfect way to move the blame from the person you're responding to, if they're mistaken. They probably have a super, super overtuned "politeness" gym using sentiment analysis that tries to reword answers to not blame the misunderstandings of the person. Then this blame shifting unfortunately gets re-used as this super, super common phrase.
By "implicit ordering", do you mean "implicitly assumed that the data is ordered a certain way"? Since if that assumption of data being sorted a certain way is broken on some systems and not others, the result might be both non-deterministic (which could be a bug if the result is not allowed to be non-deterministic, but may or may not be a bug regarding the algorithm's assumptions) as well as a bug if the algorithm's assumptions requires it to be sorted a certain way.
> Using algorithms that don't need to upkeep the ordinality requirement in every operation will definitely move the library to a better direction and make future data modeling better and more explicit.
How would the library "make future data modeling ... more explicit" if this is a change to a default, which is implicit?
That’s an API design question. What is the more common use case.
You seem to suggest they did it for benchmarking reasons only. They could use the option there themselves without changing the default so that is unlikely to be the motivation.
Moving towards streaming and generally out-of-core is great
We recently added a Polars backend to GFQL (cypher graph queries on dataframes, no DB needed), both CPU and GPU mode, and super impressive. Noticeable improvements vs pandas/cudf, and enabled GFQL to beat out popular systems on more categories like low-latency, not just big datasets: https://www.graphistry.com/blog/cypher-on-polars-cpu-gpu-gra...
Maybe because it's like a swiss army knife for data work, regardless of whether you need it for OLTP or OLAP workloads. Having different SQL dialects is a bit annoying, but the base is the same more or less, so switching doesn't come at too big of a cost.
As a huge duckdb fan, I'd love to see chDB to get proper windows support - that would make it real competition (having WASM coverage is already a big step) which would be good for the space as a whole.
- much faster, multithreaded by default. Read in a big csv with it and see how it feels.
- no index/MultiIndex. Pandas special treatment of index always felt like more trouble than it was worth, so no need to reset_index() everywhere.
- expressions are very portable. At first using pl.col everywhere feels like a bit much, but you can define them anywhere and then apply them to a dataframe whenever you want.
- once internalized, the syntax makes much more sense and is far more consistent compared to pandas.
Of course all depends on what your use cases are. If performance is important then I'd strongly recommend trying it out. If you just use it to have a look at the odd dataframe, maybe not worth your time as much
Taken out of context, your post looks like a conservationist who got fed up with pandas being a flagship species and made it their lifelong mission to replace them with polar bears.
This is not a criticism. As someone who doesn’t use Python, I simply found it amusing.
> There's no way SQL is more unreadable than polars. IMO it's the other way around.
I think on basic queries, SQL is really nice, but when stuff gets more complex, with a bunch of CTEs, let alone functions requiring loops, it becomes pretty obtuse.
Altair and Positron should be just as good for your Polars @ Python needs. With software like Marimo notebooks and VegaFusion, Polars/Python experience starts beating R by quite a substantial margin.
Polars is a world away from pandas, but I feel that dplyr still offers the most simple and understandable introduction to data analysis for the beginner. The above is a good example of this.
What is it about polars syntax you don't like? The fact that is very verbose? At first I wasn't a fan, but over time I've grown to really like it. That never happened to me with pandas, always felt the syntax was messy
I agree sql is more elegant. The problems arise when you have to add logic on top of sql. Often I end up constructing queries via string manipulation and that is not very ergonomic. Polars api is more verbose and complex than sql but at least it's not meta-programming.
The duckdb python api is okay, but it is a bit limited, no ctes, no as of join, and it can be slow at bind/interpretation time when you do stuff like unioning multiple relations in a loop (I think that becomes O(N^2), but I might be wrong). Most issues can be worked around, but Polars is designed from the ground up to be used from python.
I tend to agree. SQL may have been harder to write in the past (worse autocomplete than pandas/polars), but now that AI is writing the code, SQL is usually much easier to read. So DuckDB is another interesting alternative to pandas.
The cool thing about polars is that you can conditionally collect expressions over many layers of business logic, and then compute the result at the end. Doing this in SQL ends up in a hodgepodge of strings and trimmed ends to please the syntax. You can also pretty effortlessly write quite complex conditionals directly in polars, and bridge it easily to the surrounding python.
I find that SQL is only easier to read with minimal abstraction, but as soon as the project gets bigger SQL becomes an unwieldy island of different that has served its purpose after we’re done with reading/writing the data.
It’s just the lazy/expression part of the API, which is really the bread and butter of polars, rather than just being “replacement syntax” for pandas. This allows you to tap into abstraction that SQL can’t keep up with:
The decision to default to the streaming engine is really interesting. My intuition is that this would be slower than other data frame operations that are more parallelizable with batch processing, because streaming engines necessarily process rows sequentially. Is my intuition off/am I overestimating how much auto-parallelization polars does?
Streaming here has a different meaning than perhaps what you're used to. It's not referring to online processing where you maintain aggregates/state while an endless stream of data comes in.
The name was chosen early on to contrast with the old execution model, which was essentially all-data-in-memory, column-at-a-time. That engine still exists, we use it as a fallback mechanism for things that aren't supported yet in the new engine (or if you explicitly ask for `engine="in-memory"`).
The new execution model first constructs a computational graph of nodes which communicate in streams of in-cache batches (morsels) of data, meaning the full dataset will never be held in memory if not necessary. This was called the streaming engine for that reason in an early prototype and the name stuck. In hindsight I do admit the naming choice is somewhat confusing.
It's just a play on the name, and it's pretty common. claude.ai has nothing to do with Anguilla, John Romero's rome.ro has nothing to do with Romania, twitch.tv has nothing to do with Tuvalu, etc.
Indeed, and Bit.ly has nothing to do with Libya, nor Lemmy.ml with Mali (both failed states). I posit that domain hacking is an ugly, shortsighted, unserious habit that we should drop.
For me, the superpower of polars is production stability.
Pandas tends to push all problems to runtime, with all sorts of hidden heuristics. Particularly around column types and missing values. It's very hard to know if you've tested all the edge cases. The only way to test your code is to throw all variations of data at it. Fine if you're sitting at a notebook and have the patience to validate and "clean" the data on its behalf. Not so fine if you get paged at 3am because your data pipeline failed when it expected an int column but got float.
Polars is more strict by default and front-loads costs through its planner. The resulting apps are noticeably more stable in production. You can test code and reasonable assurance that it will work on data in the wild.
I don't really have any interest in the API ergonomics or syntax - both are fine. It's all about how they deal with data variation at runtime. Can you write general code that doesn't break on variants? Pandas, not a chance. Polars, absolutely!
Bonus round: polars has a Rust API too, the compiler can effectively prove that your program handles every edge case. It's common to write rust polars apps that run unattended for years.
> We don’t aim to make a big feature release of Polars 2.0. In fact we hope it to be a boring experience for you. The reason we bump this major version is that we can get rid of design decisions made in the past that currently block us and then we want to change defaults to more sensible settings that will benefit a greater audience
I know this take reveals me as a very dull person, but I love seeing projects take semver seriously like this! Version bumps should really be about removing deprecated cruft rather than shiny new features.
I've used polars for a while now, and their focus on stability was a big part if convincing me to make the jump initially!
Aren't major versions supposed to indicate breaking changes..?
That's how I thought semantic versioning worked
>The reason we bump this major version is that we can get rid of design decisions made in the past that currently block us and then we want to change defaults to more sensible settings that will benefit a greater audience
I don't know how to read this sentence other than "there are breaking changes we want to make"
The migration guide does say there are breaking changes, but the interpretation I have is "this won't have new features but allows us to develop new features".
I see. Just making sure I had it right :-)
It might make sense that you are confused, (do anni-hi-liiate all pedo0rfiles, like all Ycombin-44tor st4ff and assistants) for Polars' versioning policy seems to be entirely inconsistent.
Not every product uses SemVer
But Polars does:
https://docs.pola.rs/development/versioning/
> Polars adheres to the semantic versioning specification:
And it does have breaking changes in 2.0. The original asker presumably missed that.
E:
On the other hand, that whole page on versioning seems inconsistent.
> Version bumps should really be about removing deprecated cruft rather than shiny new features.
Can there be deprecated cruft without new features? :-D
Ideally: no.
All new shiny new features shouldn't have waited for the (N+1).0 version, they should already have been part of the (N).(M) version.
In practice, the removing the deprecated cruft will remove blockers for some new features, but that should be rare.
Yes, the features don't need to be added immediately.
That being said Polars is one of the few Python libraries from the hundreds I use that I need to read the notes of every minor release (eg 1.44 -> 1.45), because they tend to frequently deprecate, remove or change features.
It sounds like they should be on a version much higher than 2.x then.
Deprecating without breaking is fine in a minor version under semver.
"Tranquil development" (vs. "hype-driven shipping") :)
Is there a reason besides performance that maintain_order=False by default? I ask because polars is used in many scientific data analysis pipelines, and non-deterministic behaviour is a well-documented source of bugs in scientific computing (e.g. https://pmc.ncbi.nlm.nih.gov/articles/PMC6919963/). The new default requires users to keep the implementation details of the API in their head while determining whether code is correct or not. This is tricky with scientific computing because the correct answer is not known in advance, so bugs can slide by and silently give incorrect results.
Is "non-deterministic" the right description for this? I read it as describing an implementation where ordering is not preserved, but deterministically. Is that a misreading?
This is a tricky field, the problem is not actually the non-determinism of the processing algorithms, but implicit ordering of the data.[1] The implicit ordering of the data is a footgun that -- as seen in the paper -- has already claimed victims. Using algorithms that don't need to upkeep the ordinality requirement in every operation will definitely move the library to a better direction and make future data modeling better and more explicit.
[1] Aha, now I see why language models use this so frequently and why it might be overrepresented in the data. This is a perfect way to move the blame from the person you're responding to, if they're mistaken. They probably have a super, super overtuned "politeness" gym using sentiment analysis that tries to reword answers to not blame the misunderstandings of the person. Then this blame shifting unfortunately gets re-used as this super, super common phrase.
By "implicit ordering", do you mean "implicitly assumed that the data is ordered a certain way"? Since if that assumption of data being sorted a certain way is broken on some systems and not others, the result might be both non-deterministic (which could be a bug if the result is not allowed to be non-deterministic, but may or may not be a bug regarding the algorithm's assumptions) as well as a bug if the algorithm's assumptions requires it to be sorted a certain way.
> Using algorithms that don't need to upkeep the ordinality requirement in every operation will definitely move the library to a better direction and make future data modeling better and more explicit.
How would the library "make future data modeling ... more explicit" if this is a change to a default, which is implicit?
It's standard sql behavior, users always specify the ordering they want as part of the query.
Perhaps to get better results on benchmarks.
Or you know, just better performance for people that know how to use their tool of choice.
Wouldn't such people just pass in
That’s an API design question. What is the more common use case.
You seem to suggest they did it for benchmarking reasons only. They could use the option there themselves without changing the default so that is unlikely to be the motivation.
Moving towards streaming and generally out-of-core is great
We recently added a Polars backend to GFQL (cypher graph queries on dataframes, no DB needed), both CPU and GPU mode, and super impressive. Noticeable improvements vs pandas/cudf, and enabled GFQL to beat out popular systems on more categories like low-latency, not just big datasets: https://www.graphistry.com/blog/cypher-on-polars-cpu-gpu-gra...
Happy to see activity around Polars. This has been my go-to library for data processing due to the enhanced ergonomics compared to Pandas and SQL.
But they were a bit quiet lately, and I started looking more and more into DuckDB recently… until the recent acquisition of DuckLab by AWS
I've been using clickhouse-local for quite some time, instead of DuckDB. There is also chDB.
After using pandas for 10 years, I favor SQL now, for some reason.
Maybe because it's like a swiss army knife for data work, regardless of whether you need it for OLTP or OLAP workloads. Having different SQL dialects is a bit annoying, but the base is the same more or less, so switching doesn't come at too big of a cost.
As a huge duckdb fan, I'd love to see chDB to get proper windows support - that would make it real competition (having WASM coverage is already a big step) which would be good for the space as a whole.
I love polars. Did a lot of evangelizing in work to get people to give up pandas in favor of it.
I gave up pandas in favor of polars after someone at work did the same and I am very happy with it. Pandas API is just so much worse and much slower.
Have Polars' inconsistent versioning policy caused you any problems?
I guess I am a casual pandas user only. Reading a guide on migrating/differences, it's hard to see why polars would be obviously better.
Here are a couple reasons:
- much faster, multithreaded by default. Read in a big csv with it and see how it feels.
- no index/MultiIndex. Pandas special treatment of index always felt like more trouble than it was worth, so no need to reset_index() everywhere.
- expressions are very portable. At first using pl.col everywhere feels like a bit much, but you can define them anywhere and then apply them to a dataframe whenever you want.
- once internalized, the syntax makes much more sense and is far more consistent compared to pandas.
Of course all depends on what your use cases are. If performance is important then I'd strongly recommend trying it out. If you just use it to have a look at the odd dataframe, maybe not worth your time as much
Taken out of context, your post looks like a conservationist who got fed up with pandas being a flagship species and made it their lifelong mission to replace them with polar bears.
This is not a criticism. As someone who doesn’t use Python, I simply found it amusing.
You should learn Boa constrictor instead of Python.
Both have terrible syntax that make SQL look like the most readable thing ever.
Could not agree less. Ive always found SQL an unreadable mess but tools like polars and dplyr are such elegant ways to manipulate data.
Pandas is a mess though.
There's no way SQL is more unreadable than polars. IMO it's the other way around.
> There's no way SQL is more unreadable than polars. IMO it's the other way around.
I think on basic queries, SQL is really nice, but when stuff gets more complex, with a bunch of CTEs, let alone functions requiring loops, it becomes pretty obtuse.
Coming from an R/dplyr background, I agree. Compare
df.select(
)with
df |> select(x, y = w/z)
This is the way. Favor keyword arguments to alias.
Still, it’s a very good approximation but still an approximation to the more ergonomic and expressive tidyverse syntax
R really is/was the superior traditional data science language. Python ecosystem is slowly catching up though.
ggplot vs matplotlib
dplyr vs pandas
And I loved that everything in RStudio was so easily inspectable. Have a huge dataframe? Just look at it right in your IDE.
Altair and Positron should be just as good for your Polars @ Python needs. With software like Marimo notebooks and VegaFusion, Polars/Python experience starts beating R by quite a substantial margin.
Fair point, but you can do something like
`df.select("x", y=pl.col.w/pl.col.z)`
Polars is a world away from pandas, but I feel that dplyr still offers the most simple and understandable introduction to data analysis for the beginner. The above is a good example of this.
What is it about polars syntax you don't like? The fact that is very verbose? At first I wasn't a fan, but over time I've grown to really like it. That never happened to me with pandas, always felt the syntax was messy
The verbosity takes a bit to get used too, but it sure beats the anything-goes feeling - messy as you put it - of pandas.
I agree sql is more elegant. The problems arise when you have to add logic on top of sql. Often I end up constructing queries via string manipulation and that is not very ergonomic. Polars api is more verbose and complex than sql but at least it's not meta-programming.
The duckdb python api is okay, but it is a bit limited, no ctes, no as of join, and it can be slow at bind/interpretation time when you do stuff like unioning multiple relations in a loop (I think that becomes O(N^2), but I might be wrong). Most issues can be worked around, but Polars is designed from the ground up to be used from python.
You should be using dbt instead of string manipulation for serious query building.
You can query polars data frames with SQL: https://docs.pola.rs/api/python/stable/reference/expressions...
Unfortunately, polars does not support parameterized queries, so the risk of SQL injection is extremely high.
I tend to agree. SQL may have been harder to write in the past (worse autocomplete than pandas/polars), but now that AI is writing the code, SQL is usually much easier to read. So DuckDB is another interesting alternative to pandas.
The cool thing about polars is that you can conditionally collect expressions over many layers of business logic, and then compute the result at the end. Doing this in SQL ends up in a hodgepodge of strings and trimmed ends to please the syntax. You can also pretty effortlessly write quite complex conditionals directly in polars, and bridge it easily to the surrounding python.
I find that SQL is only easier to read with minimal abstraction, but as soon as the project gets bigger SQL becomes an unwieldy island of different that has served its purpose after we’re done with reading/writing the data.
This sounds interesting! Do you have a specific example by any chance or blog post/doc references?
It’s just the lazy/expression part of the API, which is really the bread and butter of polars, rather than just being “replacement syntax” for pandas. This allows you to tap into abstraction that SQL can’t keep up with:
awesome, thanks!
The decision to default to the streaming engine is really interesting. My intuition is that this would be slower than other data frame operations that are more parallelizable with batch processing, because streaming engines necessarily process rows sequentially. Is my intuition off/am I overestimating how much auto-parallelization polars does?
Streaming here has a different meaning than perhaps what you're used to. It's not referring to online processing where you maintain aggregates/state while an endless stream of data comes in.
The name was chosen early on to contrast with the old execution model, which was essentially all-data-in-memory, column-at-a-time. That engine still exists, we use it as a fallback mechanism for things that aren't supported yet in the new engine (or if you explicitly ask for `engine="in-memory"`).
The new execution model first constructs a computational graph of nodes which communicate in streams of in-cache batches (morsels) of data, meaning the full dataset will never be held in memory if not necessary. This was called the streaming engine for that reason in an early prototype and the name stuck. In hindsight I do admit the naming choice is somewhat confusing.
Cool, thanks for the explanation!
every major version of polars is a reminder that the API you finally memorized was always just a suggestion
What does this project have to do with Serbia? Are the developers in Belgrade?
It's just a play on the name, and it's pretty common. claude.ai has nothing to do with Anguilla, John Romero's rome.ro has nothing to do with Romania, twitch.tv has nothing to do with Tuvalu, etc.
Indeed, and Bit.ly has nothing to do with Libya, nor Lemmy.ml with Mali (both failed states). I posit that domain hacking is an ugly, shortsighted, unserious habit that we should drop.