A component of Applied Environmental Statistics this year was the creation of a video describing a moderately complex data set. This project was supported in part by the experts at the ALT Lab as well as by the donation of data sets from laboratories in Life Science and Biology.
Here is the script we went over.
The ALT Lab will be open for ALT LAB POP-UP EDIT LAB HOURS on Monday the 5th and Tuesday the 6th from 2pm-5pm. If you need some assistance completing your projects take advantage of these last times. The location of the pop-up edit lab is that open space in the 4102 suite.
Here is a short writeup on how to update R while not loosing all your packages.
There is an R and RStudio conference being offered January 11-14 in Kissimmee Florida.
Learn to write better shiny applications | Understand the new capabilities of the R Markdown authoring framework | Apply R to big data and Spark | Explore the “tidyverse” of tools for data science | Discover best practices and tips for coding with RStudio | Investigate enterprise scale development and deployment practices and tools – including the new RStudio Connect
On the Wednesday before turkey-day, we met (and when I say we, I mean the five of us), and talked about some quantitative techniques beyond that discussed in this course. They included:
Attached below is the example code for scraping water data from USGS we used as an example.
OK, here are the group assignments and a brief synopsis of the kind of data you will need to be looking for.
Members: desaixmg, turnerde2
Correlation data consists of two numerical data types. The underlying purpose of looking for correlations is to see if changes in one of the variables results in similar changes in other types of variables. However, correlation does not assume a causal function.
Members: stoskuske, nicholsonak
Linear regression is similar to correlation (perhaps more so than many people understand), wherein you have an independent variable (the predictor numerical value) and one or more dependent variables (the response numerical value). A regression model is able to be quantified producing an expected relationship among predictors in how they influence the response variable.
Members: turnerl2, atillett, tassones
Logistic regression is just like linear regression except that the response variable is dichotomous (TRUE/FALSE, PRESENCE/ABSENCE, etc).
Members: lopezr5, mccullochd, baroderique
The ANalysis Of VAriance is a way to look for differences between designated treatments (e.g., factors in R). A 1-Way anova is a way to have a single treatment (set of factors) and determine if the mean value of the response variable (a numeric variable) changes.
Members: flanarylr, hopkinsjs, bedingersd
A 2-way anova is an extension of the 1-Way that adds another treatment variable. In doing so, we can determine the extent to which either of the treatment types influence the mean of the response AND potential interactions between the predictor levels.
OK, the tallies are in!
Here are the rankings
1-10 of 14