R programming intro with examples

STATISTICAL DATA ANALYSIS USING R Additional R Programming Examples Dennis Friday 2 nd and Saturday 3 rd May, 2014. Dennis Introduction to R Statistical Programming.

Topics covered Vector and Matrix operation. File Operations. Evaluation of Probability Density Functions. Evaluation of Probability Distribution Functions, i.e. Binomial, Poisson, Normal and Uniform Distributions. Testing of Hypothesis, i.e. t, z, F, Chi-square. Linear Regression. Prediction. Sum of Squared Residuals. Residual Plots. Multiple Regression. Dennis Introduction to R Statistical Programming.

Vectors I A numeric vector is a list of numbers. The three common functions that are used to create vectors in various situations are:- c(), seq() and rep(). The function ”c()” The letter ”c” means concatenate, i.e. to join together. > c(200, 270, 250, 300, 210) # Creating a numeric vector > Saturday.sales < −c(200, 300, 320) > Sunday.sales < −c(120, 230, 200)# Saving a vector in a variable. > weekend.sales < −c(Saturday.sales, Sunday.sales)# Combining two vectors. > cities < −c(”Kerala”, ”Delhi”, ”Chennai”) # Creating a vector of characters. Dennis Introduction to R Statistical Programming.

Vectors II The functions ”cbind()”, ”rbind()” and ”seq()” > rows.sales < −cbind(Saturday.sales, Sunday.sales) # Binds Rows > columns.sales < −rbind(Saturday.sales, Sunday.sales) # Binds Columns > colnames(column.sales) < −cities # Adds titles into an array of vectors. The abbreviation ”seq” means sequence, i.e. it is used for equidistant series of numbers. > seq(4, 9) # Creates a sequence of numbers from 4 to 9 at interval of 1. # creates a sequence of numbers from 3 to 10 at interval of 2. > seq(3, 10) Dennis Introduction to R Statistical Programming.

Vectors III The function ”rep()” The abbreviation ”rep” means replicate, i.e. it is used to generate repeated values.It is used in two variants, depending on whether the second argument is a vector or a single number, for example; to repeat a series of numbers; > x = c(4, 5) > rep(x, 3) # This will repeat ”x” three times. Consider the following r-code; > rep(1 : 2, c(10, 15)) # This means repeat ”1” 10 times and ”2” 15 times. Dennis Introduction to R Statistical Programming.

Matrices I A matrix is a two-dimensional array of numbers. Example > x < −1 : 12 > dim(x) < −c(3, 4); x Alternatively, the function ”matrix()” can be used as follows; > matrix(1 : 12, nrow = 3, byrow = T) The dim assignment function sets or changes the dimension attribute of x, causing R to treat the vector of 12 numbers as a 3 x 4 matrix. Useful functions that operate on matrices include rownames(), colnames(), and the transposition function t(). Dennis Introduction to R Statistical Programming.

Matrices II Example > x < −matrix(1 : 12, nrow = 3, byrow = T) > rownames(x) < −LETTERS[1 : 3]; x > colnames(x) < −letters[1 : 3]; x # Naming rows and columns. The character vector LETTERS is a built-in variable that contains the capital letters A−Z. Similar useful vectors are letters, month.name and month.abb, which refers to lowercase letters a−z, month names, and abbreviated month names respectively. Finding the Transpose and the Inverse of a matrix. > y < −t(x); y # Gives the transpose of a matrix. > z < −solve(x); z # Gives the inverse of a matrix. Dennis Introduction to R Statistical Programming.

Matrices III Create two matrices, say, A and B. Matrix addition, subtraction and multiplication > C = c(rep(0, 3), seq(1, 6, 1), rep(seq(1, 6, 1), 2)) > A = matrix(sample(C, 12), nrow = 3, byrow = T); A > B = matrix(sample(C, 12), nrow = 3, byrow = T); B > D = A + B; D # Adds the matrix. > E = A − B; E # Subtracts matrix B form matrix A. > F = A ∗ B; F # Multiplies matrix A form matrix B element-wise. > G = A% ∗ %t(B); G # Computes matrix multiplication. Dennis Introduction to R Statistical Programming.

File operations Reading Data. The two common functions used in reading data are; scan() and read.table(), i.e. > Sales = scan(file = ”filepath”) > SALES = read.table(file = ”filepath”, header = FALSE) To merge two or more files according to the column names or row names, we use the function merge(); > AB = merge(x, y, z) # merging files. To write an output into a file, use the function write(), i.e. > output.file = write(x, file = ”file path to save the data”, ncolumns = 1, append = FALSE) Dennis Introduction to R Statistical Programming.

Some probability distributions I Normal distribution Full list and options are found in > help(Normal) command. dnorm > x < −seq(−20, 20, by = .1) > y < −dnorm(x) > plot(x, y) pnorm > x < −seq(−20, 20, by = .1) > y < −pnorm(x, mean = 3, sd = 4) > plot(x, y) Dennis Introduction to R Statistical Programming.

Some probability distributions II qnorm The next function we look at is qnorm which is the inverse of pnorm. The idea behind qnorm is that you give it a probability, and it returns the number whose cumulative distribution matches the probability. > x < −seq(0, 1, by = .05) > y < −qnorm(x) > plot(x, y) > y < −qnorm(x, mean = 3, sd = 2) > plot(x, y) rnorm > y < −rnorm(200, mean = −2) > hist(y) Dennis Introduction to R Statistical Programming.

t-distribution I dt, Help at > help(TDist) > x < −seq(−20, 20, by = .5) > y < −dt(x, df = 10) pt > x = c(−3, −4, −2, −1) > pt((mean(x) − 2)/sd(x), df = 20) qt > v < −c(0.005, .025, .05) > qt(v, df = 253) rt > rt(3, df = 10) Dennis Introduction to R Statistical Programming.

Chi-square I dchisq, Help at > help(Chisquare) > x < −seq(−20, 20, by = .5) > y < −dchisq(x, df = 10) pchisq > x = c(2, 4, 5, 6) > pchisq(x, df = 20) qchisq > v < −c(0.005, .025, .05); qchisq(v, df = 253) rchisq > rchisq(3, df = 20) Dennis Introduction to R Statistical Programming.

Testing of Hypotheses I In the one-sample t-test, we are comparing a sample mean to a known or hypothesized population mean. The null hypothesis is that the expected mean diﬀerence between the sample mean and the population mean is zero, or in other words, that the expected value of the sample mean is equal to the population mean. t-test example 1. > rnorm1 = rnorm(50, 500, 100); rnorm1 # Generate random numbers from a normal distribution with µ = 500, σ = 100. > summary(rnorm1) # Gives the descriptive summary. The mean of 517.3 is higher than µ = 500, but is it signiﬁcantly higher? To test this hypotheses, we use the function t.test(). > t.test(rnorm1, mu = 500) Dennis Introduction to R Statistical Programming.

Testing of Hypotheses II Load the in-built library datasets. t-test example 2. > library(datasets) > head(mtcars) Quiz Assuming that the data in mtcars follows the normal distribution, find the 95% confidence interval estimate of the difference between the mean gas mileage of manual and automatic transmissions. Dennis Introduction to R Statistical Programming.

Testing of Hypotheses III Solution > L = mtcars$am == 0 > mpg.auto = mtcars[L, ]$mpg; mpg.auto # automatic transmission mileage. > mpg.manual = mtcars[!L, ]$mpg; mpg.manual # manual transmission mileage. > t.test(mpg.auto, mpg.manual) Answer In mtcars, the mean mileage of automatic transmission is 17.147 mpg and the manual transmission is 24.392 mpg. The 95% conﬁdence interval of the diﬀerence in mean gas mileage is between 3.2097 and 11.2802 mpg. Dennis Introduction to R Statistical Programming.

Testing of Hypotheses IV Chi-squared test Consider the following example; > female = c(18, 102) > male = c(10, 110) > migraine = cbind(female, male); migraine > chisq.test(migraine) From the results we find that;- We determine that the chi-square test fails to reject the null hypothesis that gender and migraine susceptibility are independent. If the ratio of female to male migraine sufferers is indeed 18:6, then our result is not what we had hoped for. We may have an unusual sample or we may simply need a larger sample to obtain statistical significance. Dennis Introduction to R Statistical Programming.

Regression I Simple Linear Regression An example of grade point averages (GPAs) of 20 students along with the number of hours each student studies per week. We then use the lm() function to ﬁnd the slope and intercept terms. Hours 10 12 10 15 14 12 13 15 16 14 13 12 11 10 13 13 14 18 17 14 GPA 3.33 2.92 2.56 3.08 3.57 3.31 3.45 3.93 3.82 3.70 3.26 3.00 2.74 2.85 3.33 3.29 3.58 3.85 4.00 3.50 > results < lm(GPA ∼ Hours); summary(results) The correct interpretation of the regression equation is that a student who did not study would have an estimated GPA of 1.3728, and that for every 1-hour increase in study time, the estimated GPA would increase by 0.1489 points. Dennis Introduction to R Statistical Programming.

Regression II Example Consider the dataframe on cars, ”mtcars” in library(datasets). Test if ”mpg”-miles per gallon is aﬀected by ”disp”-displacement and ”hp”-horsepower. > myvariables = c(”mpg”, ”disp”, ”hp”) > mycars = mtcars[myvariables] > mymodel = lm(mpg ∼ disp + hp, data = mycars) > summary(mymodel) From the summary of the model, we can clearly see that the independent variables; ”disp”-Engine displacement & ”hp”-Horsepower negatively aﬀects the ”dependent variable”, i.e. for an increase in either displacement or horsepower, it will result in a decrease in the miles per gallon . Dennis Introduction to R Statistical Programming.

Regression III Another important graphical way to assess relationship between variables is by using the ”conditioning plot” as shown below. conditioning plot > coplot(mpg disp|factor(am, levels = 0 : 1, labels = c(”Automatic”, ”Manual”)), data = mycars, panel = panel.smooth, rows = 1) Regression model prediction > new = data.frame(disp = c(100, 200, 199), hp = c(121, 124, 100)) > mpg.predicted = predict(mymodel, new); mpg.predicted > data.frame(mpg.predicted, new) Dennis Introduction to R Statistical Programming.

R programming intro with examples

More Related Content

What's hot

Viewers also liked

Similar to R programming intro with examples

Recently uploaded

R programming intro with examples