APPM2720 Quiz 2

You should work on this quiz on your own and not get help from others. You are however, encouraged to use the web and any other reference materials and resources to complete these questions. All the numbered questions count equally for 10 points.

All data files are included in the APPM2720 Quiz2 directory you should also use the lectures and R code from the class as back ground and examples.

Please submit your work to me by email nychka@ucar.edu by midnight on Monday February 22, 2016. You can submit your answers:

(PART A)

The file BB200.txt contains results on the top 200 finishers from the 2014 Bolder Boulder 10K (6.2 miles) footrace. The columns are: place, names, sex/age division (M or F and age cateogry), place in the division, times for each mile, average mile time and the total time. This file can be read into R using results <- read.table("BB200.txt", header=TRUE, skip=2 ) In the file Quiz2.R is a handy function convertTime to convert the times in each column into numerical values e.g.

Mile1<- convertTime(results[,6])

Don't forget to source this function in order to use it!

(1) Explain why header=TRUE is needed. What is the effect if this is omitted?

(2) Explain why skip = 2 is needed. What is the effect if this is omitted?

(3) Convert the times to numerical values. Explain whether the first mile split time (MILE1) is predictive or not predictive of the place (PLACE). Include a figure to support your answer.

(4) Give one disadvantage or limitation of the convertTime function. (As background you should know that the Bolder Bouder has about 50,000 participants many who enjoy walking the course.)

(5) The file BB200Raw.txt is closer to the raw data that was extracted from the web. read.table does not work on this file for three separate reasons. What are the three aspects that need to be corrected for this file to be read in by read.table ?

Hint: Compare this file to the one that works!

(PART B)

The file CUNorlinQuad.jpg is an aerial photo of Norlin Quad.

(6) Read this image into R. What are the dimensions of this image and the total number of pixels?

(7) Convert this image to raster format and find the number of pixels that are exactly white. Based on the image does this answer surprise you?

Hint: Recall that the color code for pure white is "#FFFFFF" and that summing logical values (e.g. using the sum function) will count the number of TRUE cases.

(PART C)

The file testLennon.txt is a text version of the lennon image but of lower resolution so it is a smaller file.

(8) Create an image matrix using the scan and array functions and make a grey scale image plot.

(9) testLennon.txt is called a self-describing file because in the first few lines is has the information to format the image. Write an R function of the form **readImage( fname) ** that uses the scan function twice, calls the array function with the dimensions that have been read in and finally returns the image.

Your code should work for any image in this format by just giving it the file name. e.g. readImage("testLennon.txt") will return the same image matrix that you found for (7).

(PART D)

Pick a company that you are interested in and go to Yahoo Finance Type in the company name in the search window and in the company page choose the item Historical Prices in the menu on the left side.

Download prices for the past year in csv format. (Click on the option at the bottom of the table Download to Spreadsheet )

(10) Read these into R and make a plot of the daily closing and opening prices over time. Comment on any patterns.

Note: You can convert the dates in the first column to a better form by using the as.Date function. For example if the stockData is your data frame then stockData$Date <- as.Date(stockData$Date) will do it.