CompareGSDs: Compare two grain size distribution samples to see if they...

In bceaton/GSDtools: Tools for analyzing pebble counts used to measure river bed surface texture

Description

CompareGSDs is a function that takes two cumulative frequency dist- ributions with the same format as the data frames produced by MakeCFD, and uses an inverse transform approach to numerically determine whether per- centile estimates from the two samples are statistically different. The analysis requires only the cumulative frequency distribution for each sample and the number of stones measured to generate the distribution.

Usage

CompareGSDs(GSD1, GSD2, n1, n2, P = seq(5, 95, 5), alpha = 0.05)

Arguments

GSD1	is a data frame containing grain sizes (in mm) in the first col- umn, and the proportion of the distribution finer in the second column for the first sample.
GSD2	is a data frame containing grain sizes (in mm) in the first col- umn, and the proportion of the distribution finer in the second column for the second sample.
n1	is the total number of observations upon which the cumulative frequency distribution in GSD1 is based
n2	is the total number of observations upon which the cumulative frequency distribution in GSD2 is based
P	numeric vector of percentiles to be compared. The default is to compare the 5th to the 95th percentile, in increments of 5
alpha	the desired confidence level for which to calculate a confidence interval. The default is to set alpha = 0.05, for a 95% Confidence Interval

Details

The function returns a data frame listing the percentile being compared, the estimate of that percentile for sample 1, the estimate for sample 2, and a logical variable that indicates whether the percentiles are statistically different or not.

bceaton/GSDtools documentation built on June 7, 2019, 12:53 a.m.