notes.txt

Meet next monday: 2/11/2019
Get data (with HEADERS) and code
Information/details about all descriptions 

Next monday: 2/18/2019
1. get one idea from the literature he sent to try on my code (specific idea)
2. rename variables so they actually make sense and are easier to read
3. also rename the code itself so it's obvious what it means
4. read in data with column header names instead of by column index
5. clean up code so it's easier; create functions to do particular things?

Next monday: 2/25/2019
1. try plotting persistence diagrams for more variables to compare
2. non-fire data and compare to yes fire data
3. look more into tda mapper and the different functions and methods
4. find way to hover over tdamapper plot and get info about point (visnetwork?)
5. calculate betti numbers (linear algebra- rank/nullity theorem)
6. find what makes TDA unique over other methods (linear regression, etc) and a better predictor

Next Monday: 3/11/2019
1. make titles better: descriptive of what variables are being analyzed
2. play with the mapper functions (different clustering and stuff) to find more patterns
3. try to find out where the voids in the persistence diagrams are coming from (time data) and compare active/inactive data and figure out what it means
4. find some way of creating a predictor for active/inactive seasons/data
5. reduce number of plots on poster to make room to make them bigger (also make sure they're good quality) and leave room for text
6. [if you have time] find some wrapper thing to implement python code in R (betti numbers)


1. poster: add legends to plots for variables, make future work more specific, remove lambda trees/unneeded graphs not in project, 
2. comparison graph: MAYBE find graphs that will emphasize differences more, but not really needed
3. maybe look into rPython?
4. also, make table on poster in order of increasing/decreasing values 
5. note: persistence comparison with t2m and fwi (tmean_fwi and tmean_fwi_y) shows the most drastic differences between active and inactive