[R] 2 dimensional Measure of Smoothness/ Hurst Exponent questions
xiang li
lixiang128 at gmail.com
Fri Apr 8 22:58:43 CEST 2011
Dear R users:
In my research I need to come up with a method to measure the
smoothness for a closed curve (looks like ellipse, but actually not).
I was working on calculating Hurst Exponent using 2 dimensional
dataset.
I checked fArma, Randomfields and several functions in fractal
packages, but could not find a way to calculate Hurst Exponent for 2
dimensional data. For example, in fArma package, rescaled method is
defined as rsFit(x, levels = 50, minnpts = 3, cut.off = 10^c(0.7,
2.5), doplot = FALSE, trace = FALSE, title = NULL, description =
NULL). But I need to input 2 variables instead of one.
An example dataset a:
Obs Day Hour Ta Tb
121 20 0 15.72917 38.72908
122 20 1 15.22833 38.72137
123 20 2 14.58000 38.70560
124 20 3 13.92000 38.71345
125 20 4 13.54167 38.69884
126 20 5 13.71583 38.64483
127 20 6 15.16667 38.61158
128 20 7 16.92333 38.60596
129 20 8 17.81167 38.61342
130 20 9 18.76250 38.60460
131 20 10 19.84833 38.60891
132 20 11 20.53583 38.57431
133 20 12 23.00750 38.60031
134 20 13 26.49583 38.60715
135 20 14 23.42250 38.61355
136 20 15 26.01417 38.64173
137 20 16 26.76500 38.66624
138 20 17 24.54917 38.66912
139 20 18 23.67833 38.67209
140 20 19 23.20083 38.67061
141 20 20 22.12750 38.68486
142 20 21 20.42083 38.72173
143 20 22 19.25083 38.70809
144 20 23 17.44417 38.73694
You can run: plot(x=a$Ta,y=a$Tb,type="l") to see the curve. Could
someone help me to find a way to measure the smoothness of this curve?
Or could you teach me how to calculate H for 2 dimensional dataset?
Thank you so much for your help. I really appreciate!
Xiang
More information about the R-help
mailing list