So...developed an interval estimation R package, "ci". I was frustrated with finding an easy way to develop intervals when teaching confidence intervals in my basic statistics course. Really straight forward package consisting of 4 functions to estimate confidence and tolerance intervals for population means and proportions! Check it out on my GitHub! In order to download the package to your local, use the following R commands in the section below. Use the R commands help("function-name") and ?function-name to learn more about the capability of each. The package is well documented using roxygen2. *cixk**known*population standard deviation- confidence interval for population mean with**cixu***unknown*population standard deviation- tolerance interval for population proportion with**tipk***known*population proportion- confidence interval for population proportion with**cipu***unknown*population proportion
0 Comments
## Custom Boxplot in R with ggplot2After struggling the past few days to design a custom boxplot to address data from skewed distributions, I have finally figured out an interesting way to accomplish the task with "annotations" in ggplot2! Basically, needed a method to modify the "fences" of the boxplot where standard Tukey (1.5*IQR) or other multiples of the IQR do not account for skewness. Take a look below for necessary R code to construct fully customized boxplots! The following examples use the right skewed data set, 'rivers' in the R "datasets" package. Run the code in the R console to generate the custom boxplots. The standard Tukey boxplot and the modified fences boxplot by Kimber are displayed above. |
## AuthorWrite something about yourself. No need to be fancy, just an overview. ## Archives
April 2017
## Categories |