I was hoping someone might be able to help me with some statistical problems I'm having!
I am writing about Hubbert's Peak Theory and I was looking to do some stastical work on oil production figures. Basically, Hubbert's theory is that oil production will peak in some given year and then decline after that. The curve will roughly be represented by the probability density function of a logistic distribution.
If I have a dataset of yearly oil production figures, how would I go about using excel or similar package to produce a best fit curve and then test this curve for it's credentials - skewness, kurtosis, etc so that I can determine whether it really is a logistic distribution or otherwise?
I can calculate total oil production for past and future and put a figure on it - is this the same as the area under the graph showing yearly production figures? If it is, then I presume I have the value of the integral between t(start of oil production) and t(end of oil production)? Will this be useful for construction my own graph?
Is there any way for me to draw my graph alongside a 'model' gaussian curve and a model logistic distribution PDF?
Thanks EVER so much!