tunghng
commented
3 months ago Thank you for your feedback. Let's clarify:
Sample Counts Correction: The big_tech_companies.csv file contains 14 samples, not 15. It's the big_tech_stock_prices.csv that houses around 45,000 samples for in-depth analysis.
Analysis Relationship: The insights from the time series analysis in Question 1 directly inform the comparative growth study in Question 2. While Question 1 focuses on long-term trends for individual companies, Question 2 zooms in on the sector's recovery post-COVID-19 using normalized price indexes for comparison.
Variable Formulation and Visualization:
Percent Change will be calculated as ((current_price - previous_price) / previous_price) * 100. This metric will be used judiciously alongside closing prices to identify volatility and trends without misleading the analysis. The primary visualization will remain the closing prices for direct, clear comparison, while percent change could be used for specific volatility analysis. This should address the main points of your comment, focusing on accurate data representation and effective analysis methodology.
Does the 1st or the 2nd file contain 15 samples? I feel that question 2 can be answered based on the output of question 1. What are the difference between time series plot in question 1 and line chart in question 2 If you want to create additional variables, you should also describe it with a formula, i.e. how are you going to compute it
Why do you think percent change is a good choice? How are you going to visualize this info or you just compute it? Most stock price visualizations use closing price, if you use percentage change what will be your y-axis? (For instance, if stock price drops 50%, you need an increase of 100% to get back to the original price → if you use percentage change, the chart might look like price is moving upwards a lot but the fact is the price i unchanged)