Time-series

 Daniel Tafmizi

Dr. Friedman

November 4, 2024

Lis 4317

Module 10

Github: daniel.R/Work.R/LIS4370Rprog/YOYexchanges.R at main · DanielDataGit/daniel.R

My first graph is a simple time series analysis with multiple variables. This is an extension of my last assignments plots, which showed market cap over time.  This graph had the issue of not showing the changes in the smaller exchanges, as the nasdaq and nyse were about 5 times bigger than them. Market cap also does not offer much value. To fix this issue, I accumulated how the year over year percentages changed. This was fairly difficult to do. I had trouble finding resources on an efficient way to sum for each year. Eventually, I used dplyr mutate and cumsum to calculate the changes cumulatively. I used ggplot to graph it

                                 

My second graph is a time series forecast of the Euronext YOY cumulative changes. I spent some time with the forecast library but was unable to create the graph how I wanted. Below, we see a 95% and 80% confidence interval for the next 5 years. I was hoping to create a line that tracked the data more closely, not just a confidence band. I think the issue is because there is not much data for the algorithm to work with. I will need to research for one that does not require much data and can track the upward trend. I think this will be a cool addition for my final project.