Bigram Project in Java
This project contains OOP techniques such as: -Inheritance -Encapsulation -Composition
USAGE: makefile included(type make in terminal) _mainTest is included to test certain functions. 5 test files included(t1.txt .... t5.txt). ./exe filename datatype classtype • fılename: file that contains the sequence to read, there is no limit to the number of elements in the sequence. • Data type: 1 for int, 2 for strings, 3 for doubles • Class type: 1 for BigramMap , 2 for BigramDyn For example: myProg datafile 1 2 will read file name datafile that should contain integers and my program would use BigramDyn class to run all the tests.
A bigram is a specialization of N-Gram which “is a contiguous sequence of N items from a given data sequence”. For the bigram we take N=2. For example if we have a sequence of integers {1 4 6 3 7 1 4 7 2} then the number of {1 4} bigrams is 2, the number of {4 6} bigrams is 1, and the number of {3 4} bigrams is 0. Similarly, if the sequence is {qwe asd fgh sdf sdf} then the number of {qwe asd} bigrams is 0.
• readFile: takes a filename as a string parameter, reads the file, calculates all the bigrams. Throws exceptions if there are problems with opening and reading the file. • numGrams: returns the total number of bigrams calculated so far. For example, if we read an integer sequence file that contains {1 234 346 343 7234 341 434 72 234}, then numGrams() returns 8. • numOfGrams: takes two elements as bigrams and returns the number of that bigram read so far. For example, if we read an integer sequence file that contains {1 4 6 3 7 1 4 7 2}, then numOfGrams(1,4) returns 2. • operator<< prints all the bigrams and their occurences in decreasing occurent order. • maxGrams: returns the bigram that has occurred most frequently. For example, if we read an integer sequence file that contains {1 4 6 3 7 1 4 7 2}, then maxGrams() returns std::pair<int, int>(4,3).