Nice tutorials to discover R

Nice tutorials to discover R http://t.co/ckBJskmpvK via @rbloggers
— Dilir Akhtar Khan (@dilirkhan) October 27, 2013

Normalize Data in R (Calculate Z scores)

scale() function is used to create Z scores (normalize) in R.

To calculate Z score of a variable, we subtract the mean of all data points from each individual data point and divide the result by standard deviation of the variable. scale() does this in one simple call.

In R console, type

> x = c(2,4,6,8)

This creates a variable x.

To subtract the mean of the variable from each data point (this is called centering):

> scale(x, center = TRUE, scale = TRUE)  # scale = FALSE will not divide each data point by mean

> x

           [,1]
[1,] -1.1618950
[2,] -0.3872983
[3,]  0.3872983
[4,]  1.1618950
attr(,"scaled:center")
[1] 5
attr(,"scaled:scale")
[1] 2.581989

T-Test in R

98.6 t-test.xlsx   the file needs to be converted to .csv

http://ww2.coastal.edu/kingw/statistics/R-tutorials/singlesample.html

normtmp=read.csv(“e:/r/98.6 t-test.csv”,header=TRUE)
qqnorm(normtmp$tmp)
qqline(normtmp$tmp)
plot(density(normtmp$tmp))
shapiro.test(normtmp$tmp)
t.test(normtmp$tmp, mu=98.6, conf.level=.99, alternative=”two.sided”)
# output not shown

#Note: setting the alternative to “two.sided” was unnecessary, since that is the default.
We can now reject the null at any reasonable alpha level we might have chosen!
#From the sample, we might estimate the mean human body temperature to be 98.25 degrees (sample mean on the last line of output).
#A 99% CI lets us be 99% sure the population mean is between 98.08111 and 98.41735 degrees.

Different Types of Plots in R

To get the data set click this link : Friends Data from Carnegie Mellon University. data will be Data will be downloaded on your computer. Double click the downloaded file. A new session of R will start and data will be loaded in a variable named: friends.

To take a look at the data, type:
> friends

Create a table:
> t <- table(friends)

see the table:

> t

friends
No difference  Opposite sex      Same sex 
          602           434           164 

> barplot(t)

Output:

> barplot(t, horiz=T)

Try
>  barplot(t, horiz=T, main="Friends Distribution", ylab="Make Friends With", col="darkblue")

For more examples, check: http://www.statmethods.net/graphs/bar.html

Pie Chart
------------
> pie(t)

To create 3D pie chart:

> install.packages("plotrix")

>library(plotrix)

>pie3D(t, explode=.1)

Chi Square Test

Copy the following data in a text editor, add a blank line at the end and save as chisq.csv.

Heart Rate Increased, No Heart Rate Increase
Treated, 36,14
Not Treated, 30, 25

For details on the data,visit http://math.hws.edu/javamath/ryan/ChiSquare.html

What we are trying to do here is to test the effect of a drug.
Ho: The proportion of animals whose heart rate increased is independent of drug treatment.
Ha: The proportion of animals whose heart rate increased is associated with drug treatment.

Read the data into R:
> x <- read.csv("e:/r/chisq.csv")

If you didn't enter a line at the end of the file, you are likely to get the following warning:

Warning message:
In read.table(file = file, header = header, sep = sep, quote = quote,  :
incomplete final line found by readTableHeader on 'Chi_Square.csv'

However, lets run the test:

> chisq.test(x, correct=F)

Output:

        Pearson's Chi-squared test

data:  x

X-squared = 3.4177, df = 1, p-value = 0.0645

Look at the p-value. 

p-value of 0.065 is greater than the conventionally accepted of p > 0.05 we fail to reject the null hypothesis. In other words, there is no statistically significant difference in the proportion of animals whose heart rate increased.

Notes

discrete data arise from a counting process, while continuous data arise from a measuring process.

Chi square tests can only be used on actual numbers and not on percentages, proportions, means, etc.

R Video Link

http://www.twotorials.com/