The "Bootstrapped two-sample-test" gives a non-parametric two-sided "p" value (it does not assume normal distribution), based on the bootstrap posteriors for the estimate (mean or median). Inputs: Two original data sets "A" and "B", or two bootstrap posteriors from samples: "A" and "B"

This is a “R” script that describes the interquantile range test, as implemented the function interquant_r.test, is a significance test used to verify significant differences in 95% interquantile range. More specifically, the range from the 0.025 to 0.975 quantile is analysed, that encompassed 95% of the data. This test is intended to compare the precision of statistical procedures, ie. to test for diffences in 95% confidence intervals. The 95% quantile ranges (i.e., the 95% confidence interval) of two posteriors, obtained as an output from Bayesian methods or from bootstrapping (e.g., ELEFAN_Boot) are compared. A simple non-parametric statistical test (the Harrell–Davis quantile test, see Wilcox, 2012) is conducted to verify whether there are significant differences in 95% quantile ranges of posteriors, i.e., to verify whether there are significant differences in 95% confidence intervals of the original parameter estimates. It uses the function Qanova within the R package WRS2 (Mair & Wilcox, 2017), and applies it to standardized posteriors. Standardized posteriors are calculated as: standardized_data = raw_data – 0.025 quantile.

# ELEFAN_BOOT (Version 1.0.XXII.c7, beta4) # Analysis of LFD data: from RSA to confidence ellipses # Version XXII_beta4 # for LINUX_server 1 - (no plots, no beeps) # Copyright: Ralf Schwamborn, November 2017 # Only necessary code for LINUX SERVER BOOTSTRAPS (no plots, no beeps)

# ELEFAN_Perm - A permutation-based signifficance test for modal progression in LFD distributions # (Copyright: Ralf Schwamborn, April 2018) # LFD: "Length-Frequency-Distribution" # ELEFAN_Perm, version 1.0 (beta) # Null hypothesis tested: no modes, no progression (chaos,white noise) # Method: random permutation (sampling without replacement)

Performs Permutation tests (package "coin") with several taxa, one by one (loop that analyses the whole data table), generates an output table with "p" values.

Merging and Inflating two PIDs from the ZooScan equipment # v0.beta # Merges PIDs from two aliquots into one PID, simple and step-by-step # Before merging , the files are "inflated", i.e., the number of objects (data rows) in each PID file is multiplied by the Subsampling factor. # Ralf Schwamborn, 2017

## How to do a Permutation test (PERMANOVA) and a Bootstrap A Simple tutorial ## (R. Schwamborn, 2017) ### I. PERMANOVA ### II. Bootstrap

SYN-POP2 (Copyright: R. Schwamborn, 2017) is a script (i.e., a sequence of commands) for the “R” language and environment. It creates a synthetic population (“Syn.-Pop.”) of stochastically growing fish using Monte Carlo simulations. It gives length data for each group (cohort) of up to 10,000 individuals, with variable growth parameters and deterministic mortality. Estimates of growth and mortality parameters obtained with several methods can then be compared to known input values. It performs simulations with varying inputs. For each run, up to 40,000 unique individual growth trajectories, each with its own Linf and t0 and K are generated. Then, the population is submitted a posteriori to a deterministic mortality function, sampled monthly and analysed (a "Catch Curve" is plotted).

SIBMPOP4 Version 2.1. beta # (Simple: with all plots, not as function, no output table) # (For academic purposes only. SIBMPOP4 comes with absolutely no guarantee) # (R. Schwamborn, 2017) ############ # SIBMPOP4 # ############ # Age-structured, Stochastic Individual-Based Population Model (SIBM-POP4), # with well-defined cohorts and non-seasonal von Bertalanffy growth, # with fully stochastic mortality and growth. # With error terms (% C.V. of Linf, K, Z and cohort strength) # up to 11 cohorts # seasonal recruitment # adjustable duration of the recruitment season (days) # Analysis of the Catch Curve with a "modified" Powell-Wetherall plot # and output of bias (%) for Z/K and Linf estimates

AR-Wethplot2. “AR-Wethplot2” (Copyright: R. Schwamborn, 2017) is a script (i.e., a sequence of commands) for the “R” language and environment. It is an automatic routine (“A.R.”) for the analysis of length-frequency-distributions (LFDs) with the Wetherall Plot (“Wethplot”) methods in “R”. The data for regression are selected by an automatic “gamma” post-hoc selection routine. After calculation of mean lengths, it selects a specific data range from L = LFD mode+10% to L = 93%Lmax (=”gamma” selection), fits a regression model on these points and calculates the Wetherall estimates (Linf and Z/K). AR-Wethplot allows the application of the original Powell-Wetherall plot (P-W plot, Wetherall, 1986) and of the “modified” version of the method (mP-W plot, Pauly, 1986), with and without gamma selection.

IsoMC (Copyright: R. Schwamborn, 2009) is a script (i.e., a sequence of commands) for the “R” language and environment (Copyright for "R": R Foundation for Statistical Computing). It calculates probability distributions for contributions of sources A and B to a mixture, using a Monte Carlo approach. IsoMC is intended for the analysis of the shape of error propagation distributions in mixing models. 20,000 runs are used per simulation, with and without considering fractionation F. ### INSTRUCTIONS (for "R" beginners): 1. Download and install "R" (www.r-project.org), then open the IsoMC/IsoBoot textfile in any editor. 2. Write your data into "INPUT VALUES", copy the whole text and paste it into the "R" Console. ############################################# # Contact: # Ralf Schwamborn # Universidade Federal de Pernambuco (UFPE) # Recife, Brazil # e-mail: rs(at)ufpe(.)br #############################################

IsoBoot (Copyright: R. Schwamborn, 2009) is a script (i.e., a sequence of commands) for the “R” language and environment (Copyright for "R": R Foundation for Statistical Computing). It calculates a probability distribution for the contribution of a source to a mixture, using Monte Carlo simulations, and subsequent Bootstrap resampling from these simulations to estimate approximate 95% confidence intervals. IsoBoot version 1.x requires basic input parameters only, and may thus be used for any mixture of two sources, while version 3.x is specifically designed for Food Web Analysis, where inputs may include trophic level and isotopic fractionation of consumers. ### INSTRUCTIONS (for "R" beginners): 1. Download and install "R" (www.r-project.org), then open the IsoMC/IsoBoot textfile in any editor. 2. Write your data into "INPUT VALUES", copy the whole text and paste it into the "R" Console. ############################################# # Contact: # Ralf Schwamborn # Universidade Federal de Pernambuco (UFPE) # Recife, Brazil # e-mail: rs(at)ufpe(.)br #############################################

IsoBoot, v.3.1. beta (Copyright: R. Schwamborn, 2009) is a script (i.e., a sequence of commands) for the “R” language and environment (Copyright for "R": R Foundation for Statistical Computing). It calculates a probability distribution for the contribution of a source to a mixture, using Monte Carlo simulations, and subsequent Bootstrap resampling from these simulations to estimate approximate 95% confidence intervals. IsoBoot version 1.x requires basic input parameters only, and may thus be used for any mixture of two sources, while version 3.x is specifically designed for Food Web Analysis, where inputs may include trophic level and isotopic fractionation of consumers. ### INSTRUCTIONS (for "R" beginners): 1. Download and install "R" (www.r-project.org), then open the IsoMC/IsoBoot textfile in any editor. 2. Write your data into "INPUT VALUES", copy the whole text and paste it into the "R" Console. ############################################# # Contact: # Ralf Schwamborn # Universidade Federal de Pernambuco (UFPE) # Recife, Brazil # e-mail: rs(at)ufpe(.)br #############################################

IsoMC (v.2.4) IsoMC (v.2.4, Copyright: R. Schwamborn, 2009) is a script (i.e., a sequence of commands) for the “R” language and environment (Copyright for "R": R Foundation for Statistical Computing). It calculates probability distributions for contributions of sources A and B to a mixture, using a Monte Carlo approach. 20,000 runs are used per simulation, with and without considering fractionation F. ### INSTRUCTIONS (for "R" beginners): 1. Download and install "R" (www.r-project.org), then open the IsoMC textfile in any editor. 2. Write your data into "INPUTS", copy the whole text and paste it into the "R" Console. A version with spreadsheet-like inputs will be made available soon. Contact: # Ralf Schwamborn # Depto. de Zoologia # Centro de Ciências Biológicas # Universidade Federal de Pernambuco (UFPE) # 50730-540 Recife, Brazil # Phone: +55 - 81 – 2126-8859 or -7220 # e-mail: rs(at)ufpe(.)br