# subsequence_matching

**Repository Path**: su_wen_chang/subsequence_matching

## Basic Information

- **Project Name**: subsequence_matching
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MulanPSL-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2022-03-10
- **Last Updated**: 2022-03-26

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

The uploaded zip file is a clion project.
The code can be built with“cmake" and "make" instrusctions according to the "CMakeLists.txt". 

Compiling environment:
    linux compiler version : 		gcc version 5.4.0 
    windows compiler version: 		MinGW-W64-builds-4.3.5


Build index:


   0. Open file constants_and_headers.h and check the following variables，make sure you have permission to write files under these two folders: 

        const char* const INDEX_FOLDER=R"(./index_binary/);
        const char* const BINARY_SEQUENCE_PATH= R"(./index_binary/concat.binary)";

        BINARY_SEQUENCE_PATH is the  path to save the binary sequence.
        INDEX_FOLDER is the path to save the index.

   1.Clear the above two folders before buliding a new index.  If you suspect that there is a problem with the index, you could delete all the files in both folders and try to rebuild the index.

   2.check scripts/generateData.cpp, the segment length is set to {40,80,160,320} by default, this can be changed by modifying scripts/generateData.cpp and INDEX_LENGTH in constants_and_headers.h.

   3. run "make generate_data" in terminal. 
   
   4. run the obtained binary file, it wolud promote "Input the location of txt file:", input the PATH_OF_INPUT_FILE, eg:"./data10000000.txt" and wait for the index of different lengths are finished.


Run subsequence matching:
   0.Make sure the index has been built before your start.

   1.Write a script of the queries, see "in.txt" for example.
              Each line in the script is a query.
               <QUERY_TYPE> <OFFSET> <QUERY_LENGTH> <EPSILON> <ALPAH> <BETA>
	#<QUERY_TYPE> the type of query, which is one of "ED", "DTW", "CNED" and "CNDTW"
	#<OFFSET> the start position of the query sequence in the indexed series, if you want to use series from other resources, modify the "run_.cpp", find "auto Q2=fetchData(start_pos,length,BINARY_SEQUENCE_PATH);" and replace Q2 with the queries you need.
	#<QUERY_LENGTH> length of the query
	#<EPSILON> epsilon
	#<ALPAH> alpha (only for CNED and CNDTW)
	#<BETA> beta (only for CNED and CNDTW)

	Example in "in.txt":
	ED 123 40 3       
	DTW 3 100 1.1 
	CNED 1 300 3 1.1 11
	CNDTW 1 200 3 1.5 10

   2. "$make run_" 
   
   3. input the PATH_OF_SCRIPT in the terminal to run test according to the script.


Note: 
    The index building function is not well optimized and can be implemented in a more efficient way.
    You are welcomed to send emails to 19B903014@stu.hit.edu.cn for bug reporting.