Modules§
- _csv_
read_ internal polars-io - _internal
polars-io - array
polars-opsanddtype-array - binary
lazy - buffer
polars-io - byte_
source polars-ioandcloud - cat
lazyanddtype-categorical - chunkedarray
temporalTraits and utilities for temporal data. - cloud
polars-ioInterface with cloud storage through the object_store crate. - compression
polars-io - concat_
arr polars-opsanddtype-array - Data types supported by Polars.
- datetime
polars-opsandtimezones - dt
lazyandtemporal - file
polars-io - fixed_
size_ list dtype-array - function_
expr lazy - interpolate
polars-opsandinterpolate - interpolate_
by polars-opsandinterpolate_by - mode
polars-opsandmode - nan_
propagating_ aggregate polars-opsandpropagate_nans - replace
temporaland (dtype-dateordtype-datetime) - schema_
inference polars-io - series
temporal - strings
polars-ops - udf
lazy - zip
zip_with
Macros§
Structs§
- A thread-safe reference-counting pointer. ‘Arc’ stands for ‘Atomically Reference Counted’.
- Array
Name Space lazySpecialized expressions forSeriesofDataType::Array. - Represents Arrow’s metadata of a “column”.
- AsOf
Options polars-ops - Batched
CsvReader polars-io - Batched
Parquet Reader polars-io - Bounds
temporal - Bounds
Iter temporal - Brotli
Level polars-ioA valid Brotli compression level. - Specialized expressions for Categorical dtypes.
- Chained
Then lazyUtility struct for thewhen-then-otherwiseexpression. - Chained
When lazyUtility struct for thewhen-then-otherwiseexpression. - ChunkedArray
- CsvParse
Options polars-io - CsvRead
Options polars-io - CsvReader
polars-ioCreate a new DataFrame by reading a csv file. - CsvWriter
polars-ioWrite a DataFrame to csv. - CsvWriter
Options polars-ioOptions for writing CSV files. - A contiguous growable collection of
Seriesthat have the same length. - Datetime
Args lazyArguments used bydatetimein order to produce anExprof Datetime - Duration
lazy - Duration
Args lazy - Specialized expressions for modifying the name of existing expressions.
- Characterizes the name and the
DataTypeof a column. - Fields
Mapper lazy - File
Metadata polars-ioMetadata for a Parquet file. - Fixed
Size List Type dtype-array - Returned by a group_by operation on a DataFrame. This struct supports several aggregations.
- Indexes of the groups, the first index is stored separately. this make sorting fast.
- Gzip
Level polars-ioA valid Gzip compression level. - IEJoin
Options polars-ops - InProcess
Query lazy - IpcRead
Options polars-io - IpcReader
polars-ioRead Arrows IPC format into a DataFrame - IpcReader
Async polars-ioAn Arrow IPC reader implemented on top of PolarsObjectStore. - IpcScan
Options polars-io - IpcStream
Reader polars-ioRead Arrows Stream IPC format into a DataFrame - IpcStream
Writer polars-ioWrite a DataFrame to Arrow’s Streaming IPC format - IpcStream
Writer Option polars-io - IpcWriter
polars-ioWrite a DataFrame to Arrow’s IPC format - IpcWriter
Options polars-io - Join
Args lazy - Join
Builder lazy - Join
Options lazy - Json
Line Reader polars-io - Json
Reader polars-ioReads JSON in one of the formats inJsonFormatinto a DataFrame. - Json
Writer polars-ioWrites a DataFrame to JSON. - Json
Writer Options polars-io - Lazy
CsvReader lazyandcsv - Lazy
Frame lazyLazy abstraction over an eagerDataFrame. - Lazy
Group By lazyUtility struct for lazy group_by operation. - List
Name Space lazySpecialized expressions forSeriesofDataType::List. - Maps a logical type to a chunked array implementation of the physical type. This saves a lot of compiler bloat and allows us to reuse functionality.
- Just a wrapper structure which is useful for certain impl specializations.
- Null
lazyThe literal Null - Object
Type object - OptFlags
lazyAllowed optimizations. - Owned
Batched CsvReader polars-io - Owned
Object object - Parquet
Async Reader polars-ioandcloudA Parquet reader on top of the async object_store API. Only the batch reader is implemented since parquet files on cloud storage tend to be big and slow to access. - Parquet
Options polars-io - Parquet
Reader polars-ioRead Apache parquet format into a DataFrame. - Parquet
Statistics polars-ioArrow-deserialized parquet Statistics of a file - Parquet
Write Options polars-io - Parquet
Writer polars-ioWrite a DataFrame to Parquet format. - String type that inlines small strings.
- Rank
Options lazy - Rolling
Options Dynamic Window temporal - Scan
Args Ipc lazy - Scan
Args Parquet lazy - Serialize
Options polars-ioOptions to serialize logical types to CSV. - Series
- Sort options for multi-series sorting.
- Options for single series sorting.
- Special
Eq lazyWrapper type that has special equality properties depending on the inner type specialization - SplitN
Chars polars-ops - Statistics
Options polars-ioThe statistics to write - Enable the global string cache as long as the object is alive (RAII).
- Strptime
Options lazy - Struct
Array polars-ioAStructArrayis a nested [Array] with an optional validity representing multiple [Array] with the same number of rows. - Struct
Name Space lazySpecialized expressions for Struct dtypes. - Struct
Type dtype-struct - Then
lazyUtility struct for thewhen-then-otherwiseexpression. - Union
Args lazy - Unpivot
ArgsDSL lazy - Arguments for
LazyFrame::unpivotfunction - Represents a user-defined function
- When
lazyUtility struct for thewhen-then-otherwiseexpression. - Window
temporalRepresents a window in time - Zstd
Level polars-ioA valid Zstandard compression level.
Enums§
- AggExpr
lazy - The set of supported logical types in this crate.
- The time units defined in Arrow.
- Asof
Strategy polars-ops - Boolean
Function lazy - Closed
Interval polars-ops - Closed
Window temporal - A column within a
DataFrame. - Comment
Prefix polars-io - CsvEncoding
polars-io - DslPlan
lazy - Excluded
lazy - Expr
lazyExpressions that can be used in various contexts. - Function
Expr lazy - Inequality
Operator polars-ops - Interpolation
Method polars-ops - IpcCompression
polars-ioCompression codec - Join
Coalesce polars-ops - Join
Type lazy - Join
Validation lazy - Json
Format polars-ioThe format to use to write the DataFrame to JSON:Json(a JSON array) orJsonLines(each row output on a separate line). - Label
temporal - Lazy
Serde lazy - Literal
Value lazy - Maintain
Order Join polars-ops - Nested
Type lazy - Null
Strategy polars-ops - Null
Values polars-io - Operator
lazy - Parallel
Strategy polars-io - Parquet
Compression polars-ioThe compression strategy to use for writing Parquet files. - PowFunction
lazy - Quote
Style polars-ioQuote style indicating when to insert quotes around a field. - Rank
Method lazy - A dimension in a reshape.
- Search
Sorted Side polars-ops - Selector
lazy - StartBy
temporal - String
Function lazy - Struct
Function lazy - Temporal
Function lazy - Window
Mapping lazy - Window
Type lazy
Constants§
- IDX_
DTYPE Non- bigidx - NULL
lazy - URL_
ENCODE_ CHAR_ SET polars-io
Statics§
- BOOLEAN_
RE polars-io - FLOAT_
RE polars-io - FLOAT_
RE_ DECIMAL polars-io - INTEGER_
RE polars-io - POLARS_
TEMP_ DIR_ BASE_ PATH polars-io
Traits§
- Anonymous
Scan lazy - ArgAgg
polars-opsArgmin/ Argmax - AsBinary
polars-ops - AsList
polars-ops - AsString
polars-ops - Asof
Join polars-ops - Asof
Join By polars-ops - Binary
Name Space Impl polars-ops - Aggregation operations.
- Aggregations that return
Seriesof unit length. Those can be used in broadcasting operations. - Fastest way to do elementwise operations on a
ChunkedArray<T>when the operation is cheaper than branching due to null checking. - Apply kernels on the arrow array chunks in a ChunkedArray.
- Chunk
ApproxN Unique approx_unique - Cast
ChunkedArray<T>toChunkedArray<N> - Compare
SeriesandChunkedArray’s using inequality operators (<,>=, etc.) and get abooleanmask that can be used to filter rows. - Create a new ChunkedArray filled with values at that index.
- Explode/flatten a List or String Series
- Replace None values with a value
- Filter values by a boolean mask.
- Fill a ChunkedArray with one value.
- Quantile and median aggregation.
- Reverse a
ChunkedArray<T> - Chunk
Roll Apply rolling_windowThis differs from ChunkWindowCustom and ChunkWindow by not using a fold aggregator, but reusing aSerieswrapper and callingSeriesaggregators. This likely is a bit slower than ChunkWindow - Create a
ChunkedArraywith new values by index or by boolean mask. - Shift the values of a
ChunkedArrayby a number of periods. - Sort operations on
ChunkedArray. - Get unique values in a
ChunkedArray - Variance and standard deviation aggregation.
- Combine two
ChunkedArraybased on some predicate. - Chunked
Set polars-ops - Column
Binary Udf lazyA wrapper trait for any binary closureFn(Column, Column) -> PolarsResult<Column> - Columns
Udf lazyA wrapper trait for any closureFn(Vec<Series>) -> PolarsResult<Series> - Cross
Join polars-ops - Data
Frame Join Ops polars-ops - Data
Frame Ops polars-ops - Date
Methods temporal - Datetime
Methods temporal - Duration
Methods temporal - Convert
Selfinto aColumn - Used to create the tuples for a group_by operation.
- Into
Lazy lazy - Convenience for
x.into_iter().map(Into::into).collect()using aninto_vec()function. - IsFirst
Distinct is_first_distinctMask the first unique values astrue - IsLast
Distinct is_last_distinctMask the last unique values astrue - Join
Dispatch polars-ops - Reads LazyFrame from a filesystem or a cloud storage. Supports glob patterns.
- List
Name Space Impl polars-ops - Literal
lazy - MinMax
Horizontal polars-ops - Safety
- A
PolarsIteratoris an iterator over aChunkedArraywhich contains polars types. APolarsIteratormust implementExactSizeIteratorandDoubleEndedIterator. - Values need to implement this so that they can be stored into a Series and DataFrame
- Polars
Round temporal - Polars
Truncate temporal - Polars
Upsample temporal - Reinterpret
reinterpret - Rename
Alias Fn lazy - Rolling
Series polars-ops - Round
Series polars-ops - SerReader
polars-io - SerWriter
polars-io - Series
Join polars-ops - Series
Methods polars-ops - Series
OpsTime temporal - Series
Rank polars-ops - Series
Sealed polars-ops - Utility trait to slice concrete arrow arrays whilst keeping their concrete type. E.g. don’t return
Box<dyn Array>. - String
Methods temporal - String
Name Space Impl polars-ops - SumMean
Horizontal polars-ops - Take
Chunked polars-opsGather byChunkId - Take
Chunked HorPar polars-ops - Temporal
Methods temporal - Time
Methods temporal - ToDummies
polars-ops - UdfSchema
lazy
Functions§
- _coalesce_
full_ join polars-ops - _join_
suffix_ name polars-ops - Meant for internal use. In very rare conditions this can be turned off.
- abs
polars-opsConvert numerical values to their absolute value. - all
lazySelects all columns. Shorthand forcol("*"). - all_
horizontal lazyCreate a new column with the bitwise-and of the elements in each row. - any_
horizontal lazyCreate a new column with the bitwise-or of the elements in each row. - apply_
binary lazyLikemap_binary, but used in a group_by-aggregation context. - apply_
multiple lazyApply a function/closure over the groups of multiple columns. This should only be used in a group_by aggregation. - apply_
projection polars-ioand (ipcoripc_streamingorparquetoravro) - arange
lazyGenerate a range of integers. - arg_
sort_ by lazyandrangeFind the indexes that would sort these series in order of appearance. - arg_
where lazyandarg_whereGet the indices whereconditionevaluatestrue. - as_
struct lazyTake several expressions and collect them into aStructChunked. - avg
lazyFind the mean of all the values in the column namedname. Alias formean. - base_
utc_ offset temporalandtimezones - binary_
expr lazyComputeop(l, r)(or equivalentlyl op r).landrmust have types compatible with the Operator. - cast
lazyCasts the column given byExprto a different type. - clip
polars-opsSet values outside the given boundaries to the boundary value. - clip_
max polars-opsSet values above the given maximum to the maximum value. - clip_
min polars-opsSet values below the given minimum to the minimum value. - coalesce
lazyFolds the expressions from left to right keeping the first non-null values. - coalesce_
columns polars-ops - col
lazyCreate a Column Expression based on a column name. - collect_
all lazyCollect allLazyFramecomputations. - cols
lazySelect multiple columns by name. - columns_
to_ projection polars-ioand (ipcoripc_streamingoravroorparquet) - concat
lazyConcat multipleLazyFrames vertically. - concat_
arr lazyHorizontally concatenate columns into a single array-type column. - concat_
expr lazy - concat_
lf_ diagonal lazyanddiagonal_concat - Concat LazyFrames horizontally.
- concat_
list lazyConcat lists entries. - concat_
str concat_strandstringsandlazyHorizontally concat string columns in linear time - Cast null arrays to inner type and ensure that all offsets remain correct
- convert_
to_ unsigned_ index polars-ops - count_
ones polars-ops - count_
rows polars-ioRead the number of rows without parsing columns useful for count(*) queries - count_
rows_ from_ slice polars-ioRead the number of rows without parsing columns useful for count(*) queries - count_
zeros polars-ops - create_
enum_ dtype dtype-categorical - create_
sorting_ map polars-io - cum_
count polars-ops - cum_
fold_ exprs lazyanddtype-structAccumulate over multiple columns horizontally / row wise. - cum_max
polars-opsGet an array with the cumulative max computed at every element. - cum_min
polars-opsGet an array with the cumulative min computed at every element. - cum_
prod polars-opsGet an array with the cumulative product computed at every element. - cum_
reduce_ exprs lazyanddtype-structAccumulate over multiple columns horizontally / row wise. - cum_sum
polars-opsGet an array with the cumulative sum computed at every element - date_
ranges lazyandtemporalCreate a column of date ranges from astartandstopexpression. - datetime
lazyConstruct a column ofDatetimefrom the providedDatetimeArgs. - datetime_
range lazyanddtype-datetimeCreate a datetime range from astartandstopexpression. - datetime_
ranges lazyanddtype-datetimeCreate a column of datetime ranges from astartandstopexpression. - default_
join_ ids polars-ops - deserialize
polars-ioDeserializes the statistics in the column chunks from a singlerow_groupintoStatisticsassociated fromfield’s name. - diff
polars-ops - dst_
offset temporalandtimezones - dtype_
col lazySelect multiple columns by dtype. - dtype_
cols lazySelect multiple columns by dtype. - duration
lazyConstruct a column ofDurationfrom the providedDurationArgs - ensure_
duration_ matches_ dtype temporal - ensure_
is_ constant_ duration temporal - escape_
regex polars-ops - escape_
regex_ str polars-ops - expand_
paths polars-ioRecursively traverses directories and expands globs ifglobistrue. - expand_
paths_ hive polars-ioRecursively traverses directories and expands globs ifglobistrue. Returns the expanded paths and the index at which to start parsing hive partitions from the path. - expanded_
from_ single_ directory polars-ioReturnstrueifexpanded_pathswere expanded from a single directory - first
lazyFirst column in a DataFrame. - floor_
div_ series polars-ops - fold_
exprs lazyAccumulate over multiple columns horizontally / row wise. - format_
str concat_strandstringsandlazyFormat the results of an array of expressions using a format string - get_
glob_ start_ idx polars-ioGet the index of the first occurrence of a glob symbol. - get_
reader_ bytes polars-io - group_
by_ values temporalDifferent fromgroup_by_windows, where define window buckets and search which values fit that pre-defined bucket. - group_
by_ windows temporalWindow boundaries are created based on the givenWindow, which is defined by: - hor_
str_ concat polars-opsHorizontally concatenate all strings. - impl_
duration polars-ops - impl_
replace_ time_ zone polars-ops - impl_
replace_ time_ zone_ fast polars-opsIfambiguousis length-1 and not equal to “null”, we can take a slightly faster path. - in_
nanoseconds_ window temporal - index_
cols lazySelect multiple columns by index. - infer_
file_ schema polars-ioInfer the schema of a CSV file by reading through the first n rows of the file, withmax_read_rowscontrolling the maximum number of rows to read. - infer_
schema polars-ioInfers aArrowSchemafrom parquet’sFileMetadata. - int_
range lazyGenerate a range of integers. - int_
ranges lazyGenerate a range of integers for each row of the input columns. - interpolate
polars-ops - interpolate_
by polars-ops - is_
between polars-ops - is_
cloud_ url polars-ioCheck if the path is a cloud url. - is_
first_ distinct polars-ops - is_in
polars-ops - is_
last_ distinct polars-ops - is_
not_ null lazyA column which isfalsewhereverexpris null,trueelsewhere. - is_null
lazyA column which istruewhereverexpris null,falseelsewhere. - is_
positive_ idx_ uncertain polars-opsMay give false negatives because it ignores the null values. - is_
positive_ idx_ uncertain_ col polars-opsMay give false negatives because it ignores the null values. - last
lazyLast column in a DataFrame. - leading_
ones polars-ops - leading_
zeros polars-ops - len
lazyReturn the number of rows in the context. - lit
lazyCreate a Literal Expression fromL. A literal expression behaves like a column that contains a single distinct value. - map_
binary lazy - Apply a function/closure over multiple columns once the logical plan get executed.
- map_
multiple lazyApply a function/closure over multiple columns once the logical plan get executed. - materialize_
empty_ df polars-io - materialize_
projection polars-io - max
lazyFind the maximum of all the values in the column namedname. Shorthand forcol(name).max(). - mean
lazyFind the mean of all the values in the column namedname. Shorthand forcol(name).mean(). - median
lazyFind the median of all the values in the column namedname. Shorthand forcol(name).median(). - min
lazyFind the minimum of all the values in the column namedname. Shorthand forcol(name).min(). - negate
polars-ops - negate_
bitwise polars-ops - new_
int_ range polars-ops - not
lazyNegates a boolean column. - nth
lazyNth column in a DataFrame. - overwrite_
schema polars-ioandjson - Prepare the given
DslPlanfor execution on Polars Cloud. - private_
left_ join_ multiple_ keys polars-ops - quantile
lazyFind a specific quantile of all the values in the column namedname. - reduce_
exprs lazyAnalogous toIterator::reduce. - remove_
bom polars-io - repeat
lazyCreate a column of lengthncontainingncopies of the literalvalue. - repeat_
by polars-ops - replace
polars-opsReplace values by different values of the same data type. - replace_
date temporalanddtype-dateReplace specific time component of aDateChunkedwith a specified value. - replace_
datetime temporalanddtype-datetimeReplace specific time component of aDatetimeChunkedwith a specified value. - replace_
or_ default polars-opsReplace all values by different values. - replace_
strict polars-opsReplace all values by different values. - replace_
time_ zone polars-ops - resolve_
homedir polars-ioReplaces a “~” in the Path with the home directory. - split_
helper polars-ops - split_
to_ struct polars-opsanddtype-struct - str_
join polars-ops - strip_
chars polars-ops - strip_
chars_ end polars-ops - strip_
chars_ start polars-ops - strip_
prefix polars-ops - strip_
suffix polars-ops - sum
lazySum all the values in the column namedname. Shorthand forcol(name).sum(). - ternary_
expr lazy - time_
ranges lazyanddtype-timeCreate a column of time ranges from astartandstopexpression. - trailing_
ones polars-ops - trailing_
zeros polars-ops - try_
set_ sorted_ flag polars-io - unique_
counts polars-opsReturns a count of the unique values in the order of appearance. - when
lazyStart awhen-then-otherwiseexpression. - write_
partitioned_ dataset polars-ioWrite a partitioned parquet dataset. This functionality is unstable.
Type Aliases§
- AllowedOptimizations
- Array
Chunked dtype-array - An ordered sequence of
Fields - Chunk
Join OptIds polars-opsandchunked_ids - Fields
Name Mapper lazyanddtype-struct - File
Metadata Ref polars-io - GetOutput
lazy - Every group is indicated by an array where the
- IdxCa
Non- bigidx - IdxSize
Non- bigidx - IdxType
Non- bigidx - Inner
Join Ids polars-ops - Int128
Chunked dtype-i128 - Left
Join Ids polars-ops - Object
Chunked object - Opaque
Column Udf lazy - This hashmap uses an IdHasher
- Quantile
Interpol Options Deprecated - RowGroup
Iter Columns polars-io