Programming Breaks: Open Source Project: Python Implementation of Algorithms

After study of the Algorithms course by Robert Sedgewick, I have been looking for a python package that implements algorithms as outlined in his course so that i can better understand the algorithms (I like python because the language is very concise and easy to do implementation). However, while some packages in pypi have some of the implementation of Sedgewick's algorithm, none of them are quite as comprehensive and effective in implementation as Sedgewick's original Java implementation. Therefore I decided to write one on my own, which is now open sourced and available at the following link:

https://github.com/chen0040/pyalgs

Currently the python implemented algorithms cover Part I and Part II of Sedgewick's Coursera course, unit tests are provided for each algorithm so that user can understand the behavior of the algorithms, naming convention follows pypi but also mirror that of Sedgewick's Java implementation so that it is easier to follow for audience taking his course (More string manipulations and some practical examples given by Sedgewick are on their way):

More details are provided in the docs for implementation, complexities and further info.

Install pyalgs

Run the following command to install pyalgs using pip

$ pip install pyalgs

Features:

Data Structure
- Stack
  - Linked List
  - Array
- Queue
  - Linked List
  - Array
- Bag
- HashSet
- HashMap
  - Separate Chaining
  - Linear Probing
- Binary Search Tree
- Red Black Tree
- Priority Queue
  - MinPQ
  - MaxPQ
  - IndexMinPQ
- Graph
  - Simple graph
  - Edge weighted graph
  - Directed graph (digraph)
  - Directed edge weight graph
Algorithms
- Sorting
  - Selection Sort
  - Insertion Sort
  - Shell Sort
  - Merge Sort
  - Quick Sort
  - 3-Ways Quick Sort
  - Heap Sort
- Selection
  - Binary Search
- Shuffling
  - Knuth
- Union Find
  - Quick Find
  - Weighted Quick Union with path compression
Graph Algorithms
- Search
  - Depth First Search
  - Breadth First Search
- Connectivity
  - Connected Components
  - Strongly Connected Components
- Topological Sorting
  - Depth First Reverse Post Order
- Directed Cycle Detection
- Minimum Spanning Tree
  - Kruskal
  - Prim (Lazy)
  - Prim (Eager)
- Shortest Path
  - Dijkstra
  - Topological Sort (for directed acyclic graph, namely dag)
  - Bellman-Ford (for graph with negative weight as well)
- MaxFlow MinCut
  - Ford-Fulkerson
Strings
- Longest Repeated Substring
- String Sorting
  - LSD (Least Significant Digit first radix sorting)
- String Search
  - Brute force

Usage:

Data Structure

Stack

from pyalgs.data_structures.commons.stack import Stack

stack = Stack.create()
stack.push(10)
stack.push(1)

print [i for i in stack.iterate()]

print stack.is_empty()
print stack.size()

popped_item = stack.pop()
print popped_item

Queue

from pyalgs.data_structures.commons.queue import Queue

queue = Queue.create()
queue.enqueue(10)
queue.enqueue(20)
queue.enqueue(30)

print [i for i in queue.iterate()]

print queue.size()
print queue.is_empty()

deleted_item = queue.dequeue())
print deleted_item

Bag

from pyalgs.data_structures.commons.bag import Bag

bag = Bag.create()

bag.add(10)
bag.add(20)
bag.add(30)

print [i for i in bag.iterate()]

print bag.size()
print bag.is_empty()

Minimum Priority Queue

from pyalgs.data_structures.commons.priority_queue import MinPQ

pq = MinPQ.create()
pq.enqueue(10)
pq.enqueue(5)
pq.enqueue(12)
pq.enqueue(14)
pq.enqueue(2)

print pq.is_empty()
print pq.size()

print [i for i in pq.iterate()]

deleted = pq.del_min()
print(deleted)

Maximum Priority Queue

from pyalgs.data_structures.commons.priority_queue import MaxPQ

pq = MaxPQ.create()
pq.enqueue(10)
pq.enqueue(5)
pq.enqueue(12)
pq.enqueue(14)
pq.enqueue(2)

print pq.is_empty()
print pq.size()

print [i for i in pq.iterate()]

deleted = pq.del_max()
print deleted

Symbol Table using Binary Search Tree

from pyalgs.data_structures.commons.binary_search_tree import BinarySearchTree
bst = BinarySearchTree.create()

bst.put("one", 1)
bst.put("two", 2)
bst.put("three", 3)
bst.put("six", 6)
bst.put("ten", 10)

for key in bst.keys():
    print(key)

print bst.get("one"))
print bst.contains_key("two")

print bst.size()
print bst.is_empty()

bst.delete("one")

Symbol Table using Left Leaning Red Black Tree

from pyalgs.data_structures.commons.binary_search_tree import BinarySearchTree
bst = BinarySearchTree.create_red_black_tree()

bst.put("one", 1)
bst.put("two", 2)
bst.put("three", 3)
bst.put("six", 6)
bst.put("ten", 10)

print bst.get("one"))
print bst.contains_key("two")

for key in bst.keys():
    print(key)

print bst.size()
print bst.is_empty()

bst.delete("one")

Symbol Table using Hashed Map

from pyalgs.data_structures.commons.hashed_map import HashedMap
map = HashedMap.create()

map.put("one", 1)
map.put("two", 2)
map.put("three", 3)
map.put("six", 6)
map.put("ten", 10)

print map.get("one"))
print map.contains_key("two")

for key in map.keys():
    print(key)

print map.size()
print map.is_empty()

map.delete("one")

Symbol Table using Hashed Set

from pyalgs.data_structures.commons.hashed_set import HashedSet
set = HashedSet.create()

set.add("one")
set.add("two")
set.add("three")
set.add("six")
set.add("ten")

print set.contains("two")

for key in set.iterate():
    print(key)

print set.size()
print set.is_empty()

set.delete("one")

Undirected Graph

from pyalgs.data_structures.graphs.graph import Graph
def create_graph():
    G = Graph(100)

    G.add_edge(1, 2)
    G.add_edge(1, 3)

    print([i for i in G.adj(1)])
    print([i for i in G.adj(2)])
    print([i for i in G.adj(3)])

    print(G.vertex_count())
    return G

Directed Graph

from pyalgs.data_structures.graphs.graph import Digraph
def create_digraph():
    G = Digraph(100)

    G.add_edge(1, 2)
    G.add_edge(1, 3)

    print([i for i in G.adj(1)])
    print([i for i in G.adj(2)])
    print([i for i in G.adj(3)])

    print(G.vertex_count())
    return G

Edge Weighted Graph

from pyalgs.data_structures.graphs.graph import EdgeWeightGraph, Edge
def create_edge_weighted_graph():
    g = EdgeWeightedGraph(8)
    g.add_edge(Edge(0, 7, 0.16))
    g.add_edge(Edge(2, 3, 0.17))
    g.add_edge(Edge(1, 7, 0.19))
    g.add_edge(Edge(0, 2, 0.26))
    g.add_edge(Edge(5, 7, 0.28))

    print([edge for edge in G.adj(3)])

    print(G.vertex_count())
    print(', '.join([edge for edge in G.edges()]))
    return g

Directed Edge Weighted Graph

from pyalgs.data_structures.graphs.graph import DirectedEdgeWeightedGraph, Edge
def create_edge_weighted_digraph():
    g = DirectedEdgeWeightedGraph(8)

    g.add_edge(Edge(0, 1, 5.0))
    g.add_edge(Edge(0, 4, 9.0))
    g.add_edge(Edge(0, 7, 8.0))
    g.add_edge(Edge(1, 2, 12.0))
    return g

Flow Network ( for max-flow min-cut problem)

from pyalgs.data_structures.graphs.graph import FlowNetwork, FlowEdge
def create_flow_network():
g = FlowNetwork(8)
g.add_edge(FlowEdge(0, 1, 10))
g.add_edge(FlowEdge(0, 2, 5))
g.add_edge(FlowEdge(0, 3, 15))
g.add_edge(FlowEdge(1, 4, 9))
g.add_edge(FlowEdge(1, 5, 15))
g.add_edge(FlowEdge(1, 2, 4))
g.add_edge(FlowEdge(2, 5, 8))
g.add_edge(FlowEdge(2, 3, 4))
g.add_edge(FlowEdge(3, 6, 16))
g.add_edge(FlowEdge(4, 5, 15))
g.add_edge(FlowEdge(4, 7, 10))
g.add_edge(FlowEdge(5, 7, 10))
g.add_edge(FlowEdge(5, 6, 15))
g.add_edge(FlowEdge(6, 2, 6))
g.add_edge(FlowEdge(6, 7, 10))

return g

Algorithms

Union Find

from pyalgs.algorithms.commons.union_find import UnionFind

uf = UnionFind.create(10)

uf.union(1, 3)
uf.union(2, 4)
uf.union(1, 5)

print(uf.connected(1, 3))
print(uf.connected(3, 5))
print(uf.connected(1, 2))
print(uf.connected(1, 4))

Sorting

The sorting algorithms sort an array in ascending order

Selection Sort

from pyalgs.algorithms.commons.sorting import SelectionSort

a = [4, 2, 1]
SelectionSort.sort(a)
print(a)

Insertion Sort

from pyalgs.algorithms.commons.sorting import InsertionSort

a = [4, 2, 1]
InsertionSort.sort(a)
print(a)

Shell Sort

from pyalgs.algorithms.commons.sorting import ShellSort

a = [4, 2, 1, 23, 4, 5, 6, 7, 8, 9, 20, 11, 13, 34, 66]
ShellSort.sort(a)
print(a)

Merge Sort

from pyalgs.algorithms.commons.sorting import MergeSort

a = [4, 2, 1, 23, 4, 5, 6, 7, 8, 9, 20, 11, 13, 34, 66]
MergeSort.sort(a)
print(a)

Quick Sort

from pyalgs.algorithms.commons.sorting import QuickSort

a = [4, 2, 1, 23, 4, 5, 6, 7, 8, 9, 20, 11, 13, 34, 66]
QuickSort.sort(a)
print(a)

3-Ways Quick Sort

from pyalgs.algorithms.commons.sorting import ThreeWayQuickSort

a = [4, 2, 1, 23, 4, 5, 6, 7, 8, 9, 20, 11, 13, 34, 66]
ThreeWayQuickSort.sort(a)
print(a)

Heap Sort

from pyalgs.algorithms.commons.sorting import HeapSort

a = [4, 2, 1, 23, 4, 5, 6, 7, 8, 9, 20, 11, 13, 34, 66]
HeapSort.sort(a)
print(a)

Selection

Binary Selection

from pyalgs.algorithms.commons.selecting import BinarySelection
from pyalgs.algorithms.commons.util import is_sorted


a = [1, 2, 13, 22, 123]
assert is_sorted(a)
print BinarySelection.index_of(a, 13)

Shuffle

Knuth Shuffle

from pyalgs.algorithms.commons.shuffling import KnuthShuffle

a = [1, 2, 13, 22, 123]
KnuthShuffle.shuffle(a)
print(a)

Graph

Depth First Search

from pyalgs.algorithms.graphs.search import DepthFirstSearch
g = create_graph()
s = 0
dfs = DepthFirstSearch(g, s)

for v in range(1, g.vertex_count()):
    if dfs.hasPathTo(v):
        print(str(s) + ' is connected to ' + str(v))
        print('path is ' + ' => '.join([str(i) for i in dfs.pathTo(v)]))

Breadth First Search

from pyalgs.algorithms.graphs.search import BreadthFirstSearch
g = create_graph()
s = 0
dfs = BreadthFirstSearch(g, s)

for v in range(1, g.vertex_count()):
    if dfs.hasPathTo(v):
        print(str(s) + ' is connected to ' + str(v))
        print('path is ' + ' => '.join([str(i) for i in dfs.pathTo(v)]))

Connected Components

This is for undirected graph

from pyalgs.algorithms.graphs.connectivity import ConnectedComponents
G = create_graph()

cc = ConnectedComponents(G)
print('connected component count: ' + str(cc.count()))


for v in range(G.vertex_count()):
    print('id[' + str(v) + ']: ' + str(cc.id(v)))
for v in range(G.vertex_count()):
    r = randint(0, G.vertex_count()-1)
    if cc.connected(v, r):
        print(str(v) + ' is connected to ' + str(r))

Strongly Connected Components

This is for directed graph

from pyalgs.algorithms.graphs.connectivity import StronglyConnectedComponents
G = create_graph()

cc = StronglyConnectedComponents(G)
print('strongly connected component count: ' + str(cc.count()))


for v in range(G.vertex_count()):
    print('id[' + str(v) + ']: ' + str(cc.id(v)))
for v in range(G.vertex_count()):
    r = randint(0, G.vertex_count()-1)
    if cc.connected(v, r):
        print(str(v) + ' is connected to ' + str(r))

Topological Sort

from pyalgs.algorithms.graphs.topological_sort import DepthFirstOrder
G = create_graph()
topological_sort = DepthFirstOrder(G)
print(' => '.join([str(i) for i in topological_sort.postOrder()]))

Minimum Spanning Tree (Kruskal)

from pyalgs.algorithms.graphs.minimum_spanning_trees import KruskalMST
g = create_edge_weighted_graph()
mst = KruskalMST(g)

tree = mst.spanning_tree()

for e in tree:
    print(e)

Minimum Spanning Tree (Prim Lazy)

from pyalgs.algorithms.graphs.minimum_spanning_trees import LazyPrimMST
g = create_edge_weighted_graph()
mst = LazyPrimMST(g)

tree = mst.spanning_tree()

for e in tree:
    print(e)

Minimum Spanning Tree (Prim Eager)

from pyalgs.algorithms.graphs.minimum_spanning_trees import EagerPrimMST
g = create_edge_weighted_graph()
mst = EagerPrimMST(g)

tree = mst.spanning_tree()

for e in tree:
    print(e)

Directed Cycle Detection:

from pyalgs.algorithms.graphs.directed_cycle import DirectedCycle
dag = create_dag()
    dc = DirectedCycle(dag)
    assertFalse(dc.hasCycle())

Shortest Path (Dijkstra)

from pyalgs.algorithms.graphs.shortest_path import DijkstraShortestPath
g = create_edge_weighted_digraph()
s = 0
dijkstra = DijkstraShortestPath(g, s)
for v in range(1, g.vertex_count()):
    if dijkstra.hasPathTo(v):
        print(str(s) + ' is connected to ' + str(v))
        print('shortest path is ' + ' .. '.join([str(i) for i in dijkstra.shortestPathTo(v)]))
        print('path length is ' + str(dijkstra.path_length_to(v)))

Shortest Path (Topological Sort)

from pyalgs.algorithms.graphs.shortest_path import TopologicalSortShortestPath
from pyalgs.algorithms.graphs.directed_cycle import DirectedCycle
g = create_edge_weighted_digraph()
assert not DirectedCycle(g).hasCycle()
s = 0
dijkstra = TopologicalSortShortestPath(g, s)
for v in range(1, g.vertex_count()):
    if dijkstra.hasPathTo(v):
        print(str(s) + ' is connected to ' + str(v))
        print('shortest path is ' + ' .. '.join([str(i) for i in dijkstra.shortestPathTo(v)]))
        print('path length is ' + str(dijkstra.path_length_to(v)))

Shortest Path (Bellman-Ford for positive and negative edge graph)

from pyalgs.algorithms.graphs.shortest_path import BellmanFordShortestPath
from pyalgs.algorithms.graphs.directed_cycle import DirectedCycle
g = create_edge_weighted_digraph()
s = 0
dijkstra = BellmanFordShortestPath(g, s)
for v in range(1, g.vertex_count()):
    if dijkstra.hasPathTo(v):
        print(str(s) + ' is connected to ' + str(v))
        print('shortest path is ' + ' .. '.join([str(i) for i in dijkstra.shortestPathTo(v)]))
        print('path length is ' + str(dijkstra.path_length_to(v)))

MaxFlow MinCut (Ford-Fulkerson)

from pyalgs.algorithms.graphs.max_flow import FordFulkersonMaxFlow
network = create_flow_network()
ff = FordFulkersonMaxFlow(network, 0, 7)
print('max-flow: '+str(ff.max_flow_value()))

Strings

Longest Repeated Substring

from pyalgs.algorithms.strings.longest_repeated_substring import LongestRepeatedSubstringSearch
start, len = LongestRepeatedSubstringSearch.lrs('Hello World', 'World Record')
print('Hello World'[start:(start+len+1)])

Sort (LSD)

from pyalgs.algorithms.strings.sorting import LSD
LSD.sort(['good', 'cool', 'done', 'come'])

Sort (MSD)

from pyalgs.algorithms.strings.sorting import MSD
words = 'more details are provided in the docs for implementation, complexities and further info'.split(' ')
print(words)
MSD.sort(words)
print(words)

Substring Search (Brute force)

from pyalgs.algorithms.strings.substring_search import BruteForceSubstringSearch
ss = BruteForceSubstringSearch('find')
print(ss.search_in('I can find it here'))
print(ss.search_in('It is not here'))

Programming Breaks

Monday, May 1, 2017

Open Source Project: Python Implementation of Algorithms

Install pyalgs

Features:

Usage:

Data Structure

Algorithms

Graph

Strings

No comments:

Post a Comment

Blog Archive

Labels