I'm currently have a problem with interpreting the trend for 1000+ entries.
- Each entry has varying amount of periods (30~50)
- A scatterplot of the entries look like a giant blob all across the graph.
- Most of the dots is at y=0. However, the data looks unkept.
- I tried using a mean or median for each month and doing a regression. However, I don't think the result doesn't tell much.
Thus, do you think it'd be helpful if I divide the dataset into more specific category to isolate the data?