adjZipf: parameter-adjusted Zipf


adjZipf (-h -mU -kV -KW -EX -H -fY -gZ) text.spc

Parameter-adjusted extended Zipf's law. Considers four link functions (linear, logarithmic, exponential, and power) and selects the optimal model for adjustment. Adjustment by hand of the selected model parameters is possible by means of the parameters f and g. Z is estimated by requiring that V(N) . For a detailed description of the output files, see lnreZipf.

input

    text.spc: the frequency spectrum

    text.obs should be available in the working directory.

options

    -h: display on-line help

    -mU: number of frequency ranks m in the fit is set to U (default: 15)

    -kV: number of chunks for interpolation is set to V (default: 20)

    -KW: number of chunks for extrapolation is set to W (default: 20)

    -EX: extrapolation sample size is set to X (default: 2N_0)

    -H: input file has no header (default: with header)

    -fY: adjust intercept of link function by Y

    -gZ: adjust slope of link function by Z

output

    text_aZ.spc: expected frequency spectrum

    text_aZ.sp2: expected frequency spectrum at 2N_0

    text_aZ.ev2: V(N_0), E[V(N_0)] and E[V(2N_0)]

    text_aZ.int: interpolation statistics

    text_aZ.ext: extrapolation statistics

    text_aZ.sum: summary of main statistics

    text_aZ.fit: observed and expected developmental profiles of Z

    text_aZ.sta: statistics for the goodness-of-fit of the link function

         t: the t-statistic

         r: the Pearson correlation r

         a: the intercept of the (transformed) linear fit

         b: the slope of the (transformed) linear fit

         obs: linear link c(N)=a+bN

         log: logarithmic link c(N)=a + b log(N)

         exp: exponential link c(N)=e^{a + bN}

         pow: power link c(N)=e^{a + b log(N)}

[ Previous | Index | Next ]