resolve submodule problem

Author: eduardszoecs

power_ecotox

@@ -0,0 +1,21 @@
+The MIT License (MIT)
+
+Copyright (c) 2015 Eduard Szöcs
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

power_ecotox/LICENSE

@@ -0,0 +1,22 @@
+---
+title: "README"
+author: "Eduard Szöcs"
+date: "April 3, 2015"
+output: html_document
+---
+
+Power estimation for one factorial designs with count data
+===========
+
+Demo at http://52.28.43.83/shinypower/
+
+
+
+TODO:
+Add histogramm of LOECs assigned.
+Williams test switch?
+one-sided / two-sided
+
+Navbar -> choose between apps (power, population, prc, etc...)
+sidebar -> readme, 2 tabs for settings
+mainpanel -> 3 tabs for output

power_ecotox/README.Rmd

@@ -0,0 +1,3 @@
+Written by [Eduard Szöcs](http://edild.github.io/).
+
+Source Code is [here](https://github.com/EDiLD/shinypower).

power_ecotox/md/About.Rmd

@@ -0,0 +1,5 @@
+### Power Calculations and Analysis for mesocosm data
+
+Currently only power calculations (Tab `Power`) for varying experimental designs are implemented.
+
+Functions for data analysis are not implemented yet (Tabs `Community - PRC`,`Community - GLM` and `Population-GLM`).

power_ecotox/md/README_intro.md

@@ -0,0 +1,63 @@
+Power simulations for count data in one factorial designs
+===================
+
+This app runs power calculations for one factorial designs.
+One the left you can specify the expected `Effects` to simulate,
+settings for the `Simulation` design and the `Inference` procedure.
+
+To run the app with default options, simply hit the `Run simulation` Button and explore the results for a global test and a test for LOEC.
+After making changes on the left, you have to hit the `Run simulation` Button again.
+
+
+
+
+### Data
+Count data are simulated from a negative binomial distribution ($NB(\mu, \kappa)$) with mean = $\mu$ and variance = $\mu + \mu^2 / \kappa$.
+
+$\kappa$ is a dispersion parameter: 
+
+* if $\kappa \rightarrow 0$ overdispersion increases
+* if $\kappa \rightarrow \infty$ data become Poisson distributed (no overdispersion)
+
+The summary in the `Simulation-Design` tab gives mean and variances per treatment as well as graphical representation of the simulated data (assuming a sample size of 1000).
+
+### Design 
+Data are simulated for a one factorial design with 4 treatments and a step effect between treatments 2 and 3:
+
+Abundance in in treatments 1 + 2 are draw from $NB(\mu_c, \kappa)$, whereas the mean abundance in treatments 3 + 4 are reduced by factor $r$: $NB(r \times \mu_c, \kappa)$. 
+Therefore, the LOEC is at treatment 3 and NOEC at treatment 2.
+
+Note, that $\kappa$ is equal between all treatments.
+
+$\mu_c$, $\kappa$ and $r$ can be controlled by the sliders (on the left, `Effects`).
+The `Simulation-Design` tab on the right gives a graphical representation of the design.
+For demonstration purposes, 1000 data points per replicate are drawn from the specified design.
+
+
+### Models
+
+Two type of models are currently implemented:
+
+1. Linear model on transformed data
+2. Quasi-Poisson model
+
+Both are computationally feasible and give correct Type I errors.
+
+
+### Inference
+The global treatment effect is assessed using an F-test.
+LOEC is determined via Dunnett contrasts.
+The type of hypothesis (one-sided / two-sided) can be specified (on the left, `Inference`).
+
+
+### Simulations
+Power is calculated using simulations.
+The number of simulated datasets can be specified (currently max. 250 to take care of the server).
+Moreover, up to 5 sample sizes (displayed on the x-axis) can be entered. (on the left, `Simulations`)
+
+Press the `Run simulation` button to start the simulations.
+
+
+## Output
+Results for the global test and LOEC determination are given graphical and in tabular form on the right.
+Both can be downloaded using the respective buttons.