Skip to main content

Package provides Binary-, RedBlack- and AVL-Trees in Python and Cython.

Project description

Binary Tree Package

Abstract

This package provides Binary- RedBlack- and AVL-Trees written in Python and Cython/C.

This Classes are much slower than the built-in dict class, but all iterators/generators yielding data in sorted key order. Trees can be uses as drop in replacement for dicts in most cases.

Source of Algorithms

AVL- and RBTree algorithms taken from Julienne Walker: http://eternallyconfuzzled.com/jsw_home.aspx

Trees written in Python

  • BinaryTree – unbalanced binary tree

  • AVLTree – balanced AVL-Tree

  • RBTree – balanced Red-Black-Tree

Trees written with C-Functions and Cython as wrapper

  • FastBinaryTree – unbalanced binary tree

  • FastAVLTree – balanced AVL-Tree

  • FastRBTree – balanced Red-Black-Tree

All trees provides the same API, the pickle protocol is supported.

Cython-Trees have C-structs as tree-nodes and C-functions for low level operations:

  • insert

  • remove

  • get_value

  • min_item

  • max_item

  • prev_item

  • succ_item

  • floor_item

  • ceiling_item

Constructor

  • Tree() -> new empty tree;

  • Tree(mapping) -> new tree initialized from a mapping (requires only an items() method)

  • Tree(seq) -> new tree initialized from seq [(k1, v1), (k2, v2), … (kn, vn)]

Methods

  • __contains__(k) -> True if T has a key k, else False, O(log(n))

  • __delitem__(y) <==> del T[y], del[s:e], O(log(n))

  • __getitem__(y) <==> T[y], T[s:e], O(log(n))

  • __iter__() <==> iter(T)

  • __len__() <==> len(T), O(1)

  • __max__() <==> max(T), get max item (k,v) of T, O(log(n))

  • __min__() <==> min(T), get min item (k,v) of T, O(log(n))

  • __and__(other) <==> T & other, intersection

  • __or__(other) <==> T | other, union

  • __sub__(other) <==> T - other, difference

  • __xor__(other) <==> T ^ other, symmetric_difference

  • __repr__() <==> repr(T)

  • __setitem__(k, v) <==> T[k] = v, O(log(n))

  • clear() -> None, remove all items from T, O(n)

  • copy() -> a shallow copy of T, O(n*log(n))

  • discard(k) -> None, remove k from T, if k is present, O(log(n))

  • get(k[,d]) -> T[k] if k in T, else d, O(log(n))

  • is_empty() -> True if len(T) == 0, O(1)

  • items([reverse]) -> generator for (k, v) items of T, O(n)

  • keys([reverse]) -> generator for keys of T, O(n)

  • values([reverse]) -> generator for values of T, O(n)

  • pop(k[,d]) -> v, remove specified key and return the corresponding value, O(log(n))

  • pop_item() -> (k, v), remove and return some (key, value) pair as a 2-tuple, O(log(n)) (synonym popitem() exist)

  • set_default(k[,d]) -> value, T.get(k, d), also set T[k]=d if k not in T, O(log(n)) (synonym setdefault() exist)

  • update(E) -> None. Update T from dict/iterable E, O(E*log(n))

  • foreach(f, [order]) -> visit all nodes of tree (0 = ‘inorder’, -1 = ‘preorder’ or +1 = ‘postorder’) and call f(k, v) for each node, O(n)

  • iter_items(s, e[, reverse]) -> generator for (k, v) items of T for s <= key < e, O(n)

  • remove_items(keys) -> None, remove items by keys, O(n)

slicing by keys

  • item_slice(s, e[, reverse]) -> generator for (k, v) items of T for s <= key < e, O(n), synonym for iter_items(…)

  • key_slice(s, e[, reverse]) -> generator for keys of T for s <= key < e, O(n)

  • value_slice(s, e[, reverse]) -> generator for values of T for s <= key < e, O(n)

  • T[s:e] -> TreeSlice object, with keys in range s <= key < e, O(n)

  • del T[s:e] -> remove items by key slicing, for s <= key < e, O(n)

start/end parameter:

  • if ‘s’ is None or T[:e] TreeSlice/iterator starts with value of min_key();

  • if ‘e’ is None or T[s:] TreeSlice/iterator ends with value of max_key();

  • T[:] is a TreeSlice which represents the whole tree;

TreeSlice is a tree wrapper with range check and contains no references to objects, deleting objects in the associated tree also deletes the object in the TreeSlice.

  • TreeSlice[k] -> get value for key k, raises KeyError if k not exists in range s:e

  • TreeSlice[s1:e1] -> TreeSlice object, with keys in range s1 <= key < e1
    • new lower bound is max(s, s1)

    • new upper bound is min(e, e1)

TreeSlice methods:

  • items() -> generator for (k, v) items of T, O(n)

  • keys() -> generator for keys of T, O(n)

  • values() -> generator for values of T, O(n)

  • __iter__ <==> keys()

  • __repr__ <==> repr(T)

  • __contains__(key)-> True if TreeSlice has a key k, else False, O(log(n))

prev/succ operations

  • prev_item(key) -> get (k, v) pair, where k is predecessor to key, O(log(n))

  • prev_key(key) -> k, get the predecessor of key, O(log(n))

  • succ_item(key) -> get (k,v) pair as a 2-tuple, where k is successor to key, O(log(n))

  • succ_key(key) -> k, get the successor of key, O(log(n))

  • floor_item(key) -> get (k, v) pair, where k is the greatest key less than or equal to key, O(log(n))

  • floor_key(key) -> k, get the greatest key less than or equal to key, O(log(n))

  • ceiling_item(key) -> get (k, v) pair, where k is the smallest key greater than or equal to key, O(log(n))

  • ceiling_key(key) -> k, get the smallest key greater than or equal to key, O(log(n))

Heap methods

  • max_item() -> get largest (key, value) pair of T, O(log(n))

  • max_key() -> get largest key of T, O(log(n))

  • min_item() -> get smallest (key, value) pair of T, O(log(n))

  • min_key() -> get smallest key of T, O(log(n))

  • pop_min() -> (k, v), remove item with minimum key, O(log(n))

  • pop_max() -> (k, v), remove item with maximum key, O(log(n))

  • nlargest(i[,pop]) -> get list of i largest items (k, v), O(i*log(n))

  • nsmallest(i[,pop]) -> get list of i smallest items (k, v), O(i*log(n))

Set methods (using frozenset)

  • intersection(t1, t2, …) -> Tree with keys common to all trees

  • union(t1, t2, …) -> Tree with keys from either trees

  • difference(t1, t2, …) -> Tree with keys in T but not any of t1, t2, …

  • symmetric_difference(t1) -> Tree with keys in either T and t1 but not both

  • is_subset(S) -> True if every element in T is in S (synonym issubset() exist)

  • is_superset(S) -> True if every element in S is in T (synonym issuperset() exist)

  • is_disjoint(S) -> True if T has a null intersection with S (synonym isdisjoint() exist)

Classmethods

  • from_keys(S[,v]) -> New tree with keys from S and values equal to v. (synonym fromkeys() exist)

Installation

from source:

python setup.py install

or from PyPI:

pip install bintrees

Compiling the fast Trees requires Cython and on Windows is a C-Compiler necessary (MingW works fine).

Download Binaries for Windows

http://bitbucket.org/mozman/bintrees/downloads

Documentation

this README.rst

bintrees can be found on bitbucket.org at:

http://bitbucket.org/mozman/bintrees

NEWS

Version 2.0.0 June 2013

  • API change: consistent method naming with synonyms for dict/set compatibility

  • code base refactoring

  • removed tree walkers

  • removed low level node stack implementation -> caused crashes

  • optimizations for pypy: iter_items(), succ_item(), prev_item()

  • tested with CPython2.7, CPython3.3, pypy-2.0 on Win7 and Linux Mint 15 x64 (pypy-1.9)

Version 1.0.3 May 2013

  • extended iter_items(startkey=None, endkey=None, reverse=reverse) -> better performance for slicing

  • Cython implementation of iter_items() for Fast_X_Trees()

  • added key parameter reverse to itemslice(), keyslice(), valueslice()

  • tested with CPython2.7, CPython3.3, pypy-2.0

Version 1.0.2 April 2013

  • bug fix: FastRBTree data corruption on inserting existing keys

  • bug fix: union & symmetric_difference - copy all values to result tree

Version 1.0.1 February 2013

  • bug fixes

  • refactorings by graingert

  • skip useless tests for pypy

  • new license: MIT License

  • tested with CPython2.7, CPython3.2, CPython3.3, pypy-1.9, pypy-2.0-beta1

  • unified line endings to LF

  • PEP8 refactorings

  • added floor_item/key, ceiling_item/key methods, thanks to Dai Mikurube

Version 1.0.0 29.12.2011

  • bug fixes

  • status: 5 - Production/Stable

  • removed useless TreeIterator() class and T.treeiter() method.

  • patch from Max Motovilov to use Visual Studio 2008 for building C-extensions

Version 0.4.0 14.04.2011

  • API change!!!

  • full Python 3 support, also for Cython implementations

  • removed user defined compare() function - keys have to be comparable!

  • removed T.has_key(), use ‘key in T’

  • keys(), items(), values() generating ‘views’

  • removed iterkeys(), itervalues(), iteritems() methods

  • replaced index slicing by key slicing

  • removed index() and item_at()

  • repr() produces a correct representation

  • installs on systems without cython (tested with pypy)

  • new license: GNU Library or Lesser General Public License (LGPL)

Version 0.3.2 09.04.2011

  • added itemslice(startkey, endkey), keyslice(startkey, endkey), valueslice(startkey, endkey) - slicing by keys

  • tested with pypy 1.4.1, damn fast

  • Pure Python trees are working with Python 3

  • No Cython implementation for Python 3

Version 0.3.1 10.09.2010

  • runs with Python 2.7

Version 0.3.0 11.05.2010

  • low level functions written as c-module only interface to python is a cython module

  • support for the pickle protocol

Version 0.2.1 06.05.2010

  • added delslice del T[0:3] -> remove treenodes 0, 1, 2

  • added discard -> remove key without KeyError if not found

  • added heap methods: min, max, nlarges, nsmallest …

  • added Set methods -> intersection, differnce, union, …

  • added slicing: T[5:10] get items with position (not key!) 5, 6, 7, 8, 9

    T[5] get item with key! 5

  • added index: T.index(key) -> get position of item <key>

  • added item_at: T.item_at(0) -> get item at position (not key!) 0

    T.item_at(0) O(n)! <==> T.min_item() O(log(n))

Version 0.2.0 - 03.05.2010

  • TreeMixin Class as base for Python-Trees and as Mixin for Cython-Trees

Version 0.1.0 - 27.04.2010

  • Alpha status

  • Initial release

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

bintrees-2.0.0.zip (43.7 kB view details)

Uploaded Source

bintrees-2.0.0.tar.gz (32.3 kB view details)

Uploaded Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

bintrees-2.0.0.win32-py3.3.exe (308.8 kB view details)

Uploaded Source

bintrees-2.0.0.win32-py2.7.exe (312.3 kB view details)

Uploaded Source

File details

Details for the file bintrees-2.0.0.zip.

File metadata

  • Download URL: bintrees-2.0.0.zip
  • Upload date:
  • Size: 43.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No

File hashes

Hashes for bintrees-2.0.0.zip
Algorithm Hash digest
SHA256 f3263506e5ceb4c85cad58b62ee5c248e222ce2cb76155c9504cf1b18a1fc2ea
MD5 67126f0d59d7fbf5e1ff98979efe0c5d
BLAKE2b-256 8158574da5419c4da2b4d1c1c3be64010ef6d8c1f4653195173290b2bbd8ce76

See more details on using hashes here.

File details

Details for the file bintrees-2.0.0.tar.gz.

File metadata

  • Download URL: bintrees-2.0.0.tar.gz
  • Upload date:
  • Size: 32.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No

File hashes

Hashes for bintrees-2.0.0.tar.gz
Algorithm Hash digest
SHA256 3985a8e144f34e88a1c69cc57f3b419dade50cef36a3a1ca377a5c480ed2a449
MD5 a802c9f7f9204bbc591601b9172816be
BLAKE2b-256 1f9031302e4c89ac67abdde2ee497436aab66cf93c54675e4a88900e05618435

See more details on using hashes here.

File details

Details for the file bintrees-2.0.0.win32-py3.3.exe.

File metadata

File hashes

Hashes for bintrees-2.0.0.win32-py3.3.exe
Algorithm Hash digest
SHA256 f44a958bc1a8ad313b11f9b6c1b3e599c575e4787d51bf0a1d61a9f61f87ccfa
MD5 490a447509bd4f38134dcc150d306af8
BLAKE2b-256 6acf1342ec35164f74f3c7820cb66b2cc43fc7982fc593e27938cde3ce0a92ca

See more details on using hashes here.

File details

Details for the file bintrees-2.0.0.win32-py2.7.exe.

File metadata

File hashes

Hashes for bintrees-2.0.0.win32-py2.7.exe
Algorithm Hash digest
SHA256 6f5440b72b428029d30e3d80df8bbefd9d433127c2893ab2550123d687a2998c
MD5 b835233ec6fa3b7f9519b3b2627ece6d
BLAKE2b-256 a0ec07bd0c6f4eaff0523988b5513c555fe46d42ae23cdee9100afe9cbca8926

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page