Alterlab Polars

Polars

Overview

Polars is a lightning-fast DataFrame library for Python and Rust built on Apache Arrow. Work with Polars' expression-based API, lazy evaluation framework, and high-performance data manipulation capabilities for efficient data processing, pandas migration, and data pipeline optimization.

When to Use This Skill

• pandas code is too slow but the data (or the columns a lazy query touches) fits on one machine
• Building ETL or feature pipelines with lazy scans (scan_csv, scan_parquet) and query optimization
• Migrating pandas code to the Polars expression API
• Window functions, joins, and group-by aggregations over large in-memory tables

Does NOT Trigger

Quick Start

Installation and Basic Usage

Pin a recent 1.x (examples here use the Polars 1.x API):

bashCopy code
uv add 'polars>=1.0,<2'

Polars 2.0 is in release candidate (2.0.0rc2, September 2026). Its upgrade guide lists changes that alter results silently: LazyFrame.collect() defaults to the streaming engine (row order of some joins/unpivots can change), pl.concat(how="horizontal") requires equal heights (how="horizontal_extend" pads), explode() drops empty lists instead of emitting a null row, is_in refuses lossy casts, LazyFrame.profile() is removed, and headerless CSV columns are named from column_0. Check https://docs.pola.rs/releases/upgrade/ before moving to 2.x.

Basic DataFrame creation and operations:

pythonCopy code
import polars as pl

# Create DataFrame
df = pl.DataFrame({
    "name": ["Alice", "Bob", "Charlie"],
    "age": [25, 30, 35],
    "city": ["NY", "LA", "SF"]
})

# Select columns
df.select("name", "age")

# Filter rows
df.filter(pl.col("age") > 25)

# Add computed columns
df.with_columns(
    age_plus_10=pl.col("age") + 10
)

Core Concepts

Expressions

Expressions are the fundamental building blocks of Polars operations. They describe transformations on data and can be composed, reused, and optimized.

Key principles:

• Use pl.col("column_name") to reference columns
• Chain methods to build complex transformations
• Expressions are lazy and only execute within contexts (select, with_columns, filter, group_by)

Example:

pythonCopy code
# Expression-based computation
df.select(
    pl.col("name"),
    (pl.col("age") * 12).alias("age_in_months")
)

Lazy vs Eager Evaluation

Eager (DataFrame): Operations execute immediately

pythonCopy code
df = pl.read_csv("file.csv")  # Reads immediately
result = df.filter(pl.col("age") > 25)  # Executes immediately

Lazy (LazyFrame): Operations build a query plan, optimized before execution

pythonCopy code
lf = pl.scan_csv("file.csv")  # Doesn't read yet
result = lf.filter(pl.col("age") > 25).select("name", "age")
df = result.collect()  # Now executes optimized query

When to use lazy:

• Working with large datasets
• Complex query pipelines
• When only some columns/rows are needed
• Performance is critical

Benefits of lazy evaluation:

• Automatic query optimization
• Predicate pushdown
• Projection pushdown
• Parallel execution

For detailed concepts, load references/core_concepts.md.

Common Operations

Select

Select and manipulate columns:

pythonCopy code
# Select specific columns
df.select("name", "age")

# Select with expressions
df.select(
    pl.col("name"),
    (pl.col("age") * 2).alias("double_age")
)

# Select all columns matching a pattern
df.select(pl.col("^.*_id$"))

Filter

Filter rows by conditions:

pythonCopy code
# Single condition
df.filter(pl.col("age") > 25)

# Multiple conditions (cleaner than using &)
df.filter(
    pl.col("age") > 25,
    pl.col("city") == "NY"
)

# Complex conditions
df.filter(
    (pl.col("age") > 25) | (pl.col("city") == "LA")
)

With Columns

Add or modify columns while preserving existing ones:

pythonCopy code
# Add new columns
df.with_columns(
    age_plus_10=pl.col("age") + 10,
    name_upper=pl.col("name").str.to_uppercase()
)

# Parallel computation (all columns computed in parallel)
df.with_columns(
    pl.col("value") * 10,
    pl.col("value") * 100,
)

Group By and Aggregations

Group data and compute aggregations:

pythonCopy code
# Basic grouping
df.group_by("city").agg(
    pl.col("age").mean().alias("avg_age"),
    pl.len().alias("count")
)

# Multiple group keys
df.group_by("city", "department").agg(
    pl.col("salary").sum()
)

# Conditional aggregations
df.group_by("city").agg(
    (pl.col("age") > 30).sum().alias("over_30")
)

For detailed operation patterns, load references/operations.md.

Aggregations and Window Functions

Aggregation Functions

Common aggregations within group_by context:

• pl.len() - count rows
• pl.col("x").sum() - sum values
• pl.col("x").mean() - average
• pl.col("x").min() / pl.col("x").max() - extremes
• pl.first() / pl.last() - first/last values

Window Functions with over()

Apply aggregations while preserving row count:

pythonCopy code
# Add group statistics to each row
df.with_columns(
    avg_age_by_city=pl.col("age").mean().over("city"),
    rank_in_city=pl.col("salary").rank().over("city")
)

# Multiple grouping columns
df.with_columns(
    group_avg=pl.col("value").mean().over("category", "region")
)

Mapping strategies:

• group_to_rows (default): Preserves original row order
• explode: Changes the row count (use in select, not with_columns); output is grouped
• join: Creates list columns

Data I/O

Supported Formats

Polars supports reading and writing:

• CSV, Parquet, JSON, Excel
• Databases (via connectors)
• Cloud storage (S3, Azure, GCS)
• Google BigQuery
• Multiple/partitioned files

Common I/O Operations

CSV:

pythonCopy code
# Eager
df = pl.read_csv("file.csv")
df.write_csv("output.csv")

# Lazy (preferred for large files)
lf = pl.scan_csv("file.csv")
result = lf.filter(...).select(...).collect()

Parquet (recommended for performance):

pythonCopy code
df = pl.read_parquet("file.parquet")
df.write_parquet("output.parquet")

JSON:

pythonCopy code
df = pl.read_json("file.json")
df.write_json("output.json")

For comprehensive I/O documentation, load references/io_guide.md.

Transformations

Joins

Combine DataFrames:

pythonCopy code
# Inner join
df1.join(df2, on="id", how="inner")

# Left join
df1.join(df2, on="id", how="left")

# Join on different column names
df1.join(df2, left_on="user_id", right_on="id")

Concatenation

Stack DataFrames:

pythonCopy code
# Vertical (stack rows)
pl.concat([df1, df2], how="vertical")

# Horizontal (add columns)
pl.concat([df1, df2], how="horizontal")

# Diagonal (union with different schemas)
pl.concat([df1, df2], how="diagonal")

Pivot and Unpivot

Reshape data:

pythonCopy code
# Pivot (wide format): on= is the column whose values become new columns
df.pivot("product", index="date", values="sales")

# Unpivot (long format)
df.unpivot(index="id", on=["col1", "col2"])

For detailed transformation examples, load references/transformations.md.

Pandas Migration

Polars offers significant performance improvements over pandas with a cleaner API. Key differences:

Conceptual Differences

• No index: Polars uses integer positions only
• Strict typing: No silent type conversions
• Lazy evaluation: Available via LazyFrame
• Parallel by default: Operations parallelized automatically

Common Operation Mappings

Key Syntax Patterns

Pandas sequential (slow):

pythonCopy code
df.assign(
    col_a=lambda df_: df_.value * 10,
    col_b=lambda df_: df_.value * 100
)

Polars parallel (fast):

pythonCopy code
df.with_columns(
    col_a=pl.col("value") * 10,
    col_b=pl.col("value") * 100,
)

For comprehensive migration guide, load references/pandas_migration.md.

Best Practices

Polars 1.x API notes (renamed since 0.x)

These older names appear in stale tutorials and LLM training data; use the 1.x form:

• read_csv(dtypes=...) -> read_csv(schema_overrides=...)
• pl.Utf8 -> pl.String
• pl.NUMERIC_DTYPES -> import polars.selectors as cs; cs.numeric()
• join(how="outer") -> join(how="full") (and how="full" no longer coalesces keys; pass coalesce=True for old behavior)
• pivot(columns=...) -> pivot(on=...) (on is the first positional arg)
• collect(streaming=True) -> collect(engine="streaming")
• read_database(connection_uri=...) -> read_database_uri(uri=...)

Performance Optimization

1. Use lazy evaluation for large datasets:

   pythonCopy code
   lf = pl.scan_csv("large.csv")  # Don't use read_csv
   result = lf.filter(...).select(...).collect()
   

2. Avoid Python functions in hot paths:

   • Stay within expression API for parallelization
   • Use .map_elements() only when necessary
   • Prefer native Polars operations

3. Use the streaming engine to lower peak memory:

   pythonCopy code
   lf.collect(engine="streaming")  # `streaming=True` is deprecated
   

4. Select only needed columns early:

   pythonCopy code
   # Good: Select columns early
   lf.select("col1", "col2").filter(...)
   
   # Bad: Filter on all columns first
   lf.filter(...).select("col1", "col2")
   

5. Use appropriate data types:

   • Categorical for low-cardinality strings
   • Appropriate integer sizes (i32 vs i64)
   • Date types for temporal data

Expression Patterns

Conditional operations:

pythonCopy code
pl.when(condition).then(value).otherwise(other_value)

Column operations across multiple columns:

pythonCopy code
df.select(pl.col("^.*_value$") * 2)  # Regex pattern

Null handling:

pythonCopy code
pl.col("x").fill_null(0)
pl.col("x").is_null()
pl.col("x").drop_nulls()

For additional best practices and patterns, load references/best_practices.md.

Resources

This skill includes comprehensive reference documentation:

references/

• core_concepts.md - Detailed explanations of expressions, lazy evaluation, and type system
• operations.md - Comprehensive guide to all common operations with examples
• pandas_migration.md - Complete migration guide from pandas to Polars
• io_guide.md - Data I/O operations for all supported formats
• transformations.md - Joins, concatenation, pivots, and reshaping operations
• best_practices.md - Performance optimization tips and common patterns

Load these references as needed when users require detailed information about specific topics.