Inputting data into SAS

I. We’ll first input some information for a sample of technology firms. The variables inputted are, in order, revenues (in millions), revenue growth, return on equity, total shareholder return and profits (in millions). The comma delimited data is techa.csv.

Notes:

a. Always create a data statement before performing such tasks as reading in or manipulating data. You can do a number of things within an individual data step but you probably would want to separate your tasks into different data steps.

b. The infile statement directs SAS to look for an external data file at the specified location.

c. The delimiter=’,’ statement indicates to SAS that the variables are separated by commas. SAS can input various data formats.

d. The input statement specifies the names of the variables inputted. The number of variables should normally equal the number of columns of the data.

e. The proc print statement prints out the data on screen. Notice there are a few missing values.

II. Create a SAS program to read in the data set techb.csv. The data set has two variables: revenues and profits. Use the proc print statement to look at what you have read in.

III. Create a SAS program to read in the data set techc.csv. The data set has three variables: revenue growth, return on equity and total shareholder return. Use the proc print statement to look at what you have read in. Do you notice the missing values?

IV. Now we will take a data set and make it more easily readable by SAS. The data is the original Excel spreadsheet the above variables came from: tech2.xls. Eliminate the title on top and the variable names. Save as comma delimited. Keep track of what the variables are. To input a non-numeric variable into SAS, add the $ sign within the input statement after the nonnumeric variable name.

V. We next take this comma delimited data set tech.csv.

a. Create a variable indicating whether the company originates from the US or not.

b. Calculate the mean and standard deviation in shareholder return for US and non-US firms separately.

c. Calculate the same statistics for shareholder return but now distinguish between firms whose revenue growth was at least ten percent and those with growth below ten percent.

VI. This file is already formatted to be read by SAS. The data represents characteristics of a sample of cities. The variables in order: city name, percent non-hispanic white, percent over age 65, percent of adults with college degree, percent employed in professional occupation, median household income.

a. Input data.

b. Calculate the proportion of cities with a median income over $60,000.

c. Calculate the mean percent employed in professional occupations separately for cities with median income over $60,000 and less than (or equal to) $60,000.

VII. Download and set up the following excel data set to be read in SAS:

Library Data. Perform the following regression.

y – total library expenditure (expend)

X₁ – number of residents (residents)

X₂ – dummy variable for city/county run library (citlib)

X₃ – percent school-aged children (school)

X₄ – median income (medinc)