Skip to content

nitinkarolla/Search-SimiliarityJoin-AdvanceDataManagement

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

README for Assignment 1

Requirements

Implement SimilarityJoinED(). You can add your own auxiliary function in simJoin.py

Read and Write functions are provided in main.py. Please DO NOT modify main.py (But you might want to change the input filenames while testing.)

Submission

Submit your simJoin.py via Canvas We will use our own main.py and grading script to test your code.

Note for grade and test

When you finish your implementation, running python main.py should gives your output as a csv file (out.csv by default). Example testsets and their outputs with corresponding threshold tau are provided in ./testset/.

Be careful about the datatype in each column (ID and E-distance are all int, do not print something looks like 0.0, 1.0, etc.) as well as output format.

We will grade using diff in Linux.

Example: diff sample_test1.csv out.csv

Final large test Set

We will use a large dataset with less than 1 million strings, each string with length at least 10 characters and at most 200 characters for final evaluation.


SimilarityJoinED()

  1. Number of Parameters: 2

    • dat: A list generated from input file (each element is a line in the input file)
    • threshold: tau
  2. Input data format

    • one string per line
    • ID is not recorded in the file, but please see the line index as the ID for each string. Note: START FROM 0!
  3. Output

    • You should output an csv file WITHOUT HEADER, dimension is n by 3, where n is the number of combination that fits for the threshold requirements. Since output function has already provided, you should generate a list of list of 3 tuple<ID1,ID2,Edistance> (all of them are int) in Python and pass it to WriteResults()
    • Note about order of the tuples: Lexicographical order of <ID1, ID2>. Ascendant to ID1; If ID1 is the same, ascendant to ID2.

Example: self-joint with threshold = 1:

  • Run
python main.py
  • Input file sample_test1.txt (4 lines, with index 0 ,1, 2, 3)
hello
hell
hella
hallo
  • Expected output out.csv
0,1,1
0,2,1
0,3,1
1,2,1

Output Format Note

  1. Contains 3 columns <ID1,ID2,Edistance>, separated with ,.
  2. Do not write headers and index

About

Implementations of Pass Join near similarity search in Python.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages