OK, I have re-run the data using your correction.
It was 237 deaths in Uigi 3 days ago and 277 today. (If I have that wrong still, someone correct me).
That corresponds to exponential growth and an e-folding time of 19 days.
We are seeing data that range from 9 days to 108 days.
However, we can conclude the following:
1. It is still growing exponentially.
2. It is obviously not contained.
3. The exponential time constant is probably not as fast as 9 days, nor as slow as 100, and is beginning to converge to a nominal number of roughly 20 days.
The approximate growth (I'll wait before I replot) is:
May 1: 550 cumulative cases
June 1: 2,605 cumulative cases
July 1: 12,272 cumulative cases
I doubt if one exponential curve would fit all the data, particularly as people change their behavior in response to the disease (i.e., reducing contacts with others) or if efforts at containment by medical staff or government were even partially successful. Changes like those would affect the number of transmissions to others and slow the growth of cases.
Still, the modeling you are doing is very useful, incomplete data or not. Thank you.
I found this on recominomics.com commentary:
In spite of removal of cases outside of Uige, the total number of reported Marburg cases to date continues to rise. The number of reported cases remaining on the official tally has risen to 264, which is 16 shy of the record death toll for Ebola in Republic of Congo in 1976. Since the beginning of the outbreak last fall, the number of monthly cases has doubled each month. Last month there were approximately 70 cases that have remained on the official list. Since there have been approximately 130 new cases this month that have not been discarded, it seems likely that the number of cases in April will be more than double the number of cases in March.
http://www.recombinomics.com/News/04200503/Marburg_Toll_264.html