iCAMP (version 1.5.12)

icamp.boot: Bootstrapping analysis of icamp results

Description

Use bootstrapping to estimate the variation of relative importance of each process in each group, and compare the difference between groups.

Usage

icamp.boot(icamp.result, treat, rand.time = 1000, compare = TRUE,
           silent = FALSE, between.group = FALSE, ST.estimation = FALSE)

Arguments

icamp.result

data.frame object, from the output of icamp.big. the first two columns are sample IDs, the third to seventh columns are the relative importance of the five ecological processes.

treat

matrix or data.frame, a one-column (n x 1) matrix indicating the group or treatment of each sample, rownames are sample IDs. if input a n x m matrix, only the first column is used.

rand.time

integer, bootstrapping times. default is 1000.

compare

logic, whether to compare icamp reults between different groups.

silent

logic, if FALSE, some messages will show during calculation.

between.group

logic, whether to analyze between-treatment turnovers.

ST.estimation

logic, whether to estimate stochasticity as the total relative importance of dispersal and drift.

Value

Output is a list with three elements.

summary

data.frame, summary of each group.

Group: group name from the input "treat".

Process: process name from the icamp.result.

Observed: the mean relative importance of each process in each group.

Mean, Stdev, Min, Quartile25, Median, Quartile75, and Max: mean, standard deviation, minimum, 25 percent-quantile, median, 75 percent-quantile, and maximum of bootstrapping results, respectively.

Lower.whisker, Lower.hinge, Mediean.1, Higher.hinge, Higher.whisker, Outerlier1...: boxplot elements.

compare

data.frame, summary of comaprison between each two groups. First two columns are group names. From the third column, different indexes for comparison are showed, including Cohen's d (Cohen.d), effect size magnitude according to Cohen's d (Effect.Size), and P value from bootstrapping test (P.value).

boot.detail

a list of matrixes, each matrix corresponds to a group, showing detailed bootstrapping results in each random draw.

Details

Bootstrapping is implemented by random draw samples with replacement, to estimate the variation of relative importance of each process in each group, and calculate the relative difference, effect size, and significance of the difference between each two groups.

References

Ning, D., Yuan, M., Wu, L., Zhang, Y., Guo, X., Zhou, X. et al. (2020). A quantitative framework reveals ecological drivers of grassland microbial community assembly in response to warming. Nature Communications, 11, 4717.

See Also

icamp.big

Examples

# NOT RUN {
data("icamp.out")
data("example.data")
treatment=example.data$treat
rand.time=20 # usually use 1000 for real data.
icampbt=icamp.boot(icamp.result = icamp.out$bNRIiRCa, treat = treatment,
                   rand.time = rand.time, compare = TRUE,
                   between.group = TRUE, ST.estimation = TRUE)
# }