I am working with an ebola dataset from Guinea, 2014:(http://currents.plos.org/outbreaks/article/estimating-the-reproduction-number-of-zaire-ebolavirus-ebov-during-the-2014-outbreak-in-west-africa/).

My main objective is to use this dataset to test an epidemic model by eventually fitting it to this set. The main problem I am having is that I would like to track the epidemic by days starting from 22-march, to August-20 (151 days), however, my set contains only 58 points.

Is it possible to interpolate these points reasonably well by keeping the points associated with each date and approximating the rest?

I have exported the data, but even just plotting cases and deaths vs time is resulting in an error message. Here's what I tried:

data = Take[Import["Ebola-Guinea.csv", 
"DateStringFormat" -> {"Day", " ", "MonthNameShort", " ", 
  "Year"}]] // TableForm


However, this results in:

<<1>> is not a valid dataset or list of datasets. >>

It is important for me to keep this time scale as I will be using it to estimate parameters. Would someone care to offer some insight as to how to approach this problem? Many thanks.

  • 1
    $\begingroup$ Try removing // TableForm (and your use of Take[...] is redundant too) but without a sample of your data it's difficult to answer. $\endgroup$ Dec 4 '15 at 15:33
  • $\begingroup$ As I understand, the "deaths" column is total deaths to the date, while the cases column is strange. For instance, why are there sometimes less cases then on previous days? Such as 14th and 17th July? Anyway, as far as deaths go, if these are cumulative to that date, then a linear interpolation and subsequent application of Floor seems ok here. $\endgroup$
    – LLlAMnYP
    Dec 4 '15 at 15:36
  • $\begingroup$ @ LLIAMnYP: as I understand it, the total number of cases is the sum of infected and dead. The number of cases would decline when infected move to the recovered compartment, or die. $\endgroup$
    – Lucif3r
    Dec 4 '15 at 15:53
  • $\begingroup$ The @ ping apparently doesn't work if you leave a space between @ and the username. If it's the sum of infected and dead, then death is not a cause for decline, though recovery is. However if it's just infected, then I'd expect deaths to grow faster. However what happened between 8th and 12th July (and some others), that deaths decreased? In any case, disregarding these oddities, my suggestion about taking the floor still stands. $\endgroup$
    – LLlAMnYP
    Dec 4 '15 at 16:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.