search

joshuaye1024 / covidApp

covid-data
0 stars 0 forks source link

Add additional Analytic Features #1

Closed leafdad closed 3 years ago

leafdad commented 3 years ago

  1. Add moving average statistics using pandas.rolling(period) function. Allow user to select the period, like 5 days or 7 days, etc.

  2. Perform statistics on the changes in positive cases, hospitalization, and death. You should try different lag paramters, such regress posistive cases against 1 week delay of hospitalization, and 2-week delay for number of death etc. The goal is for us to get a sense how cases evolves in death....

  3. On the side, would be good to investigate other Python GUI than TK....

joshuaye1024 commented 3 years ago
  1. use OLS module to write regression option, use lag option as parameter to check where r^2 is maximized
leafdad commented 3 years ago

  1. Specifically, for a given rolling period d (say 7 days), and a lag time (say 15 days) between changes in d-rolling cases vs changes in the lagged d-rolling deaths, you calculate the resuling R^2 using statsmodels OLS function, with the changes in cases as the independent variable and the changes in deaths as the dependent variable. Do this for each reasonable combination of the rolling period d and time lag, and find the one with the highest R^2. Of course, d and lag have to make intuitive sense, which is where your judgment comes in.

  2. You can do the same for hospitslization, and ICU as well in place of deaths. Also find the lag relationships between hopspitalization/ICU and deaths as well, and inutitively, the lag is much shorter.

  3. If you finish the above, and still want to have time for more challenges, you could attempt to find the relationship between returns (changes in log prices for the rolling period d) on ZOOM stock price and positive cases, with or withour lag....