Primary Types
Penguin organizes data into PTable
s. Data is stored in columnar format inside PTypedColumn
s,
however to make it easier to work with, PTable
‘s are composed of multiple PColumn
s, which
hold within them a PTypedColumn
.
Penguin methods only throw errors of type PError
, which makes them easy to catch and handle.
-
A collection of named
PColumns
.A PTable, also known as a data frame, represents a tabular collection of data.
Invariants:
- Each column must have the same number of elements.
- Column names are unique.
Declaration
Swift
public struct PTable
extension PTable: CustomStringConvertible
extension PTable: Equatable
-
A dtype-erased column of data.
See moreDeclaration
Swift
public struct PColumn
extension PColumn: Equatable
-
Undocumented
See moreDeclaration
Swift
@dynamicMemberLookup public struct PTypedColumn<T> where T : PCSVParsible, T : PDefaultInit, T : PStringParsible, T : Comparable, T : Hashable
extension PTypedColumn: Equatable
extension PTypedColumn: CustomStringConvertible
-
Undocumented
See moreDeclaration
Swift
public class Aggregation
-
PIndexSet represents a (non-strict) subset of indices of a column or table.
PIndexSet is used for masking and other operations on a
PTypedColumn
, aPColumn
, and aPTable
. APIndexSet
is most often created via operations on the column types, such asPColumn
‘snils
property, which returns aPIndexSet
representing all the indices (rows) containing nils.To help catch errors, operations on
PIndexSet
s check to ensure they represent collections of the same size. e.g. When aPIndexSet
is used to select rows out of aPTable
, thecount
property of thePIndexSet
is checked to ensure it is exactly equal to thePTable
’scount
property.
See morePIndexSet
supports both in-place and chaining set operations.Declaration
Swift
public struct PIndexSet : Equatable
-
Undocumented
See moreDeclaration
Swift
public enum PError : Error, Equatable
extension PError: CustomStringConvertible