Project 1 Name: Fang Zhou 1. Structure of the project Root |-cleanr.py Cleaner class. It's used to clean the data. |-parser.py Parser class. It's used to parse the data. |-reuter.py Reuter class. The Reuter class to hold a lot of info. |-vector.py Main funciton. |-env_bash.sh Environment script for bash |-env_tcsh.sh Environment script for tcsh |-Makefile |-readme 2. Set Python environment Please run the env script (env_bash.sh or env_tcsh.sh) before you run the program. Use 'source env_bash.sh' or 'source env_bash.sh' to run it. It depends on the shell you are using. 3. How to run You should give the program three input arguments. The first one is the input filename. The second one is the output filename. Note: Because we have three feature vectors, so once you give the output filename, the program uses vc0, vc1, and vc2 as the extension of the outputfile. So you will get three output files with the same prefix. The third one is the number of attributes. You can set the number of values as you want. For example: python -O ./vector.py data result 100 This function will generate three files: result.vc1, result.vc2, and result.vc3, for feature vectors based on TF-IDF, word frequency, and general information, respectively. Note: I also prepare a Makefile for the grader. If you run 'make', it will run the program with the whole dataset and generate three vector files name result.vc1, result.vc2, resut.vc3. It may take 2-3 mins on our deparment cluster. 4. How to review You can use vim or other editors to open the output files. The first line in each output file should contain the attribute names. So you can easily know there are how many attributes and what the meanings of them are. The other lines show the values of the feature vectors in the unit of document, which means one line represents one document in this file. If you do not want the attribute names, you can use awk, sed and any commands you like to remove the first line. If you have any problems, please let me know. Email: zhou.1250@osu.edu