0

From my understanding, log measures relative change, which is the same as what percentages do? Is my understanding faulty? If so, please guide me to the correct one. This is the formula I used to calculate percentage increase in cases-

percent_increase_cases = ((new_cases - new_cases_one_day_before)/new_cases_one_day_before) * 100

log percent change

Also, should log scale be used on total cases or new cases?

callmeanythingyouwant
  • 115
  • 4
  • 1
    [This](https://stats.stackexchange.com/a/74927/7071) might clear up some misconception. – dimitriy Sep 24 '20 at 05:16
  • You missed a step: the bottom graph will, at least for small absolute values, approximate a graph of the *ratios* of successive values on a log scale -- but that's not what you're plotting in the top graph. The two graphs aren't comparable. – whuber Sep 24 '20 at 13:50
  • I watched the logarithm video on khan academy and it cleared out some doubts. My basics are weak and I'll get the hang of them the more log problems I solve from some textbook probably. Thanks. – callmeanythingyouwant Sep 25 '20 at 14:01

0 Answers0