StrainRanking (version 1.2)

gmcpic.test: Function implementing the Generalized Monte Carlo plug-in test with calibration (GMCPIC test)

Description

The GMCPIC test is a procedure to test the equality of the vectors of probabilities of two multinomial draws. The test statistics that is used is the multinomial-density statistic.

Usage

gmcpic.test(x, B, M, weights, threshold)

Arguments

x

[2-column matrix] Column 1 (resp. 2) contains the vector of observed frequencies in population 1 (resp. 2).

B

[Integer] Number of Monte Carlo simulations.

M

[Integer] Number of repetitions for the calibration.

weights

[Numeric] Vector of weights in [0,1] that are tried for the calibration.

threshold

[Numeric] Targeted risk level of the test; value in [0,1].

Value

list with INPUT arguments (x, B, M, weights and threshold) and the following items:

calibrated.weight

Weight selected by the calibration procedure.

p.value

Test p-value.

reject.null.hypothesis

Logical indicating whether the null hypothesis is rejected or not at the risk level specified by threshold.

Message

Details about the p-value interpretation.

Details

The GMCPIC test was developed to test the similarity of two pathogen compositions based on small samples and sparse data.

References

Soubeyrand S, Garreta V, Monteil C, Suffert F, Goyeau H, Berder J, Moinard J, Fournier E, Tharreau D, Morris C, Sache I (2017). Testing differences between pathogen compositions with small samples and sparse data. Phytopathology 107: 1199-1208. http://doi.org/10.1094/PHYTO-02-17-0070-FI

Examples

Run this code
# NOT RUN {
## Load Pathogen Compositions of M. oryzae collected in Madagascar
data(PathogenCompositionMoryzaeMadagascar)
x=t(PathogenCompositionMoryzaeMadagascar)

## Apply the GMCPIC test (use B=10^3, M=10^4 to get a robust result)
# }
# NOT RUN {
testMada=gmcpic.test(x, B=10^2, M=10^3, weights=seq(0.5,0.99,by=0.01),threshold=0.05)
# }
# NOT RUN {
testMada
# }
# NOT RUN {
## Apply the Chi-squared test
chisq.test(x, simulate.p.value = TRUE, B = 10000)
# }

Run the code above in your browser using DataCamp Workspace