Analysis of Scientific Data Summary of R functions This sheet gives a summary of the functions in R that you will meet in the course this semester. You can also get more information using help() or ? in R. Importing Data Use data = read. csv("file") for CSV data and data = read. delim("file") for tab-separated data. Use View(data), str(data), names(data), nrow(data) or head(data) as checks that your data was imported correctly. Use data$X, for example, to extract the variable X from the data frame data. Use data$Z = data$Y1 - data$Y2, for example, to calculate a new variable. Use subset (data, X == "x") to extract a new data frame with just only the cases where X is x. Use data$X = ordered(data$X, c(A, B, C) ) to make X an ordinal variable with A < B < C, for example. Lattice Graphics Use library(lattice) to load the lattice graphics library. Strip plots stripplot(data$Y); stripplot(~ Y, data); stripplot(Y ~ X, data) Histogram histogram(data$Y); histogram(~ Y, data) Density plot densityplot (data$Y); densityplot(~ Y, data) Box plot bwplot(data$Y); bwplot(~ Y, data); bwplot(Y ~ X, data) Quantile plot qqmath(data$Y); qqmath(~ Y, data) Scatter plot xyplot(Y ~ X, data); use type="p" for data points, "1" for joining with lines, "g" for a grid, "r " for a regression line, and "smooth" for a smoothing line. Bar chart barchart(table(data$x, data$Y) ) Spine plot spineplot (table(data$x, data$Y) ) Use the group=Z option to separate by variable Z. Add a title to your plot using the main="title" option. Change the axes labels using xlab and ylab. Get a simple legend with auto. key=TRUE. Summary Statistics Get basic statistics with summary (data), mean(data$Y), median(data$Y), sd(data$Y), IQR(data$Y), and fivenum(data$Y). For categorical data use table(data$Y) or table(data$Y, data$X) for a two-way table. prop. table() can be applied to table () to get proportions and marginal proportions. Use aggregate( ) to get statistics by group. For example, aggregate(Y ~ X, data, mean) gives the mean Y value for each category in X. This outputs a data frame which you can also use for plotting.
Basic Inference One-sample t test t.test(data$x) t.test(Y~X, data) Two-sample t test See power. t. test() for power calculations One proportion prop.test(x, n) Two proportions prop.test(table(data$x, data$Y) ) Chi square test chisq. test(table(data$x, data$Y) ) Model Building For each of the following you can use functions like summary(), anova(), predict(), and residuals() for details of the analysis. Linear regression 1m(Y ~ X, data) ANOVA aov(Y ~ X, data) Logistic regression glm(Y ~ X, data, family="binomial") Use Y ~ X1*X2 fora two-factor model with an interaction term or Y ~ X1+X2 for a model without an interaction term. Other Calculations Logarithms log(x) for natural logs and log10 (x) for base 10 logs Exponentials exp(x) for ex and 10^x for 10x Factorial factorial (n) gives n! Combinations choose(n, k) Distributions Note that the following distribution functions all give areas to the left (matching the theoretical definitions of these functions) whereas the tables in the textbook all give areas to