rowante wrote:It really blows my mind that that article was published the year I was born.
Web Hubble Telescope, did you get in contact with the author, is he still around?
Name: Denny Payton
E-Mail:
Date/Time: 3/30/2004 1:06 PM
Message:
When George Pazik’s name came up recently, a couple fellows shared their appreciation for him. I had the opportunity to meet him over the phone yesterday, and was saddened to discover a weak 83-year-old man who was unable to put into words, everything he wanted to say. He’s had a long struggle with medical problems, and spent four years in a nursing home, I think just getting out recently. He said there’d been a period of time when he couldn’t walk, but that he can now, with the assistance of a walker. When I told him he was the one who introduced me to Buck Perry by an article in Fishing Facts, I could see that it really encouraged him.
Anyways, I have his mailing address, so if anyone would like to send him a note of encouragement or appreciation, please email me [email protected] and I’ll give it to you. I don’t think I should post it here. He could sure use it.
Thanks,
Denny Payton
Auburn, IN
chrispi wrote:I have found that using the Hubbert distribution to find the future price of oil yields three possibilities, if both human demand and oil supply follow Hubbert's distribution:I used the Mathematica program and its statistical packages. What do you think?
- The rate of population growth is equal to the rate of oil extraction. Then the price of oil follows an S shaped curve bounded above and below by finite limits, with the inflection point right around Peak Oil.
- The rate of population growth is greater than the rate of oil extraction. Then the price of oil follows a bell curve, getting ever more expensive until after Peak Oil, then getting cheaper again.
- The rate of population growth is less than the rate of oil extraction. Then the price of oil will fall exponentially until we reach Peak Oil, whereupon the price will rise exponentially.
chrispi wrote:It's a simple model, I know, but sometimes simplicity is best when capturing the essence of price behavior.
If you look at the historical chart for oil prices you'll find that they went down exponentially for most of the twentieth century, ignoring price spikes like the Iran Crisis and the like. This agrees with the last model, where pop growth < oil extraction, which suggests that the price of oil will skyrocket in the coming decades, ignoring price spikes and collapses, of course.
EnergySpin wrote:chrispi wrote:It's a simple model, I know, but sometimes simplicity is best when capturing the essence of price behavior.
If you look at the historical chart for oil prices you'll find that they went down exponentially for most of the twentieth century, ignoring price spikes like the Iran Crisis and the like. This agrees with the last model, where pop growth < oil extraction, which suggests that the price of oil will skyrocket in the coming decades, ignoring price spikes and collapses, of course.
Check the Verhulst thread .. we have done the exercise
How did you do the fitting with Mathematica?
I used Mathematica for the same application (- humans) ... hope you understand Nonlinear regression
chrispi wrote:EnergySpin wrote:chrispi wrote:It's a simple model, I know, but sometimes simplicity is best when capturing the essence of price behavior.
If you look at the historical chart for oil prices you'll find that they went down exponentially for most of the twentieth century, ignoring price spikes like the Iran Crisis and the like. This agrees with the last model, where pop growth < oil extraction, which suggests that the price of oil will skyrocket in the coming decades, ignoring price spikes and collapses, of course.
Check the Verhulst thread .. we have done the exercise
How did you do the fitting with Mathematica?
I used Mathematica for the same application (- humans) ... hope you understand Nonlinear regression
Yeah Mathematica is awesome wrt statistical curve fitting. I used it on historical prices to find the correct logistic model (exponential decline followed by skyrocketing prices) which I obtained by graphing a simple division:
Plot[PDF[LogisticDistribution[humanpeak,humancoeff],t]/PDF[LogisticDistribution[oilpeak,oilcoeff],t],{t,tmin,tmax}]
The humanpeak, humancoeff, oilpeak, oilcoeff parameters were found using a nonlinear fit on historical data. The tmin and tmax parameters are of your choosing. Happy datamining!
EnergySpin wrote:chrispi wrote:EnergySpin wrote:chrispi wrote:It's a simple model, I know, but sometimes simplicity is best when capturing the essence of price behavior.
If you look at the historical chart for oil prices you'll find that they went down exponentially for most of the twentieth century, ignoring price spikes like the Iran Crisis and the like. This agrees with the last model, where pop growth < oil extraction, which suggests that the price of oil will skyrocket in the coming decades, ignoring price spikes and collapses, of course.
Check the Verhulst thread .. we have done the exercise
How did you do the fitting with Mathematica?
I used Mathematica for the same application (- humans) ... hope you understand Nonlinear regression
Yeah Mathematica is awesome wrt statistical curve fitting. I used it on historical prices to find the correct logistic model (exponential decline followed by skyrocketing prices) which I obtained by graphing a simple division:
Plot[PDF[LogisticDistribution[humanpeak,humancoeff],t]/PDF[LogisticDistribution[oilpeak,oilcoeff],t],{t,tmin,tmax}]
The humanpeak, humancoeff, oilpeak, oilcoeff parameters were found using a nonlinear fit on historical data. The tmin and tmax parameters are of your choosing. Happy datamining!
Be careful with the Levenberg Marquadt and check the curvature measures of nonlinearity. Curve fitting these equations can be bitch
chrispi wrote:EnergySpin wrote:chrispi wrote:EnergySpin wrote:chrispi wrote:It's a simple model, I know, but sometimes simplicity is best when capturing the essence of price behavior.
If you look at the historical chart for oil prices you'll find that they went down exponentially for most of the twentieth century, ignoring price spikes like the Iran Crisis and the like. This agrees with the last model, where pop growth < oil extraction, which suggests that the price of oil will skyrocket in the coming decades, ignoring price spikes and collapses, of course.
Check the Verhulst thread .. we have done the exercise
How did you do the fitting with Mathematica?
I used Mathematica for the same application (- humans) ... hope you understand Nonlinear regression
Yeah Mathematica is awesome wrt statistical curve fitting. I used it on historical prices to find the correct logistic model (exponential decline followed by skyrocketing prices) which I obtained by graphing a simple division:
Plot[PDF[LogisticDistribution[humanpeak,humancoeff],t]/PDF[LogisticDistribution[oilpeak,oilcoeff],t],{t,tmin,tmax}]
The humanpeak, humancoeff, oilpeak, oilcoeff parameters were found using a nonlinear fit on historical data. The tmin and tmax parameters are of your choosing. Happy datamining!
Be careful with the Levenberg Marquadt and check the curvature measures of nonlinearity. Curve fitting these equations can be bitch
Are you talking about skewness? (∫(x-m)^3 f(x)dx)
doufus wrote:What data? What source? What reliability? What effect does changing
technology have in the data? Can that even be modelled?
Users browsing this forum: No registered users and 10 guests