This project aims to analyse stocks from different sectors.

Please refer the project report and codes in the files.

File CrawlYahoo.java contains code to crawl yahoo finance data. By changing url in linktoscrape to any yahoo finance stock url, data can be crawled. Format.java contains code to separate and extract only the required data. Pass a crawled data file.

To run the python codes

python3 stock_gev.py ( to perform the qq plot, chi square test) python3 stock_pe.py ( to plot the predicted vs measured stock) extremogram.R (to check the extreme dependency) gev.m( to calculate the gev parameters)

Note: Code for background work for EVT and GARCH are not uploaded.

HClust.R contains code for hierarchical clustering. import Dec6 or Dec1to8 data and run the script.

NMF Clustering:

Toy data- runnable code. Dataset “datamatrix”, which is the dataset for a week’s data(dec1st to dec 8th) Dataset “namematrix” which consists of the names for the companies.
Open the program in Rstudio. Click on run.

Contributors: Krupa Hegde, Ankit Jain