UCLA subscribes to the database of economic variables provided by CITICORP Database Services. These are typically time-series data, running over different time periods, and they are updated several times a year. Not every variable you might be interested in using will be contained here, but there are a couple of thousand.
The reason we are covering this is because this database can potentially save a lot of time in the process of acquiring data for some future empirical term paper you may attempt in some other field course at UCLA, or in your future studies. Most institutions have CITIBASE data, and many firms subscribe as well. Knowledge about how to access and download data from CITIBASE is another potentially valuable job skill.
1. In the lab, select MS-DOS. Then get a look at the different online databases available via SSC by typing the ONLINE command. Some of these databases are contained on permanently mounted CD-ROM drives; others need to be requested, since they are not used as frequently. Databases with a dollar sign next to them require that you sign up for a debit account with SSC. A couple of dollars goes quite a long way. [Check on current status of this information.]
2. Type CITIBASE to access the free Citibase data bank. When asked to identify a drive, enter a: if you have a diskette with you, or c:\ if you wish to use the main directory of the hard drive on the machine you are using during this session.
3. The CITIBASE menu has four options:
1 Run Citibase
2 Browse CITIMAXI.LST
3 Print list of CITI1- CITI5 contents
4 Exit
Choose 2 to explore the range of time-series variables available in the aggregate Citibase data bank. Be sure to WRITE DOWN the acronym for the variable(s) you wish to retrieve from the database. For example, the CCIURR variable (Consumer installment credit outstanding: revolving, retailers, million dollars, EOM, Not Seasonally Adjusted). Exit from the browser using the ESC key.
4. Now choose option 1 Run Citibase from the CITIBASE menu. Note that the Data Drive is Q:, the Data Path is \CITIBASE, and the filename depends upon what series you wish to retrieve. In the current edition, there are five separate files necessary to hold all of the series in the data set. They are:
CITI1 series up to GAW86Our example series, CCIURR, will thus be found in the file called CITI1. Type in the Data Drive, Data Path, and filename information.
CITI2 series up to HSBSOU
CITI3 series up to LPGOVS
CITI4 series up to PAWFA
CITI5 series up to WTR
5. You will now see a menu that includes:
F1 VIEWChoose VIEW to begin with, because we need some more information before we attempt to download any variables. Page down using the arrow keys until you get to the CCIURR variable. Hit ENTER. Study the description of the series at the top of the page. Check its frequency (e.g. monthly) and note its START DATE yrmo and its END DATE yrmo. (For example, yrmo might be 8001, meaning 1980, January). Record the available observations carefully, since you will be asked for them later in this process. If you are happy with this variable, hit ESC until you get back to the menu where F2 RETRIEVE is an option.
F2 RETRIEVE
F10 QUIT
6. When you opt to RETRIEVE, you will again have to fill in the Data Drive, Data Path, and filename information as in step 4. You will then be presented with a TEMPLATE menu, superimposed by a four-option menu as follows:
CREATEChoose EXPORT. When given a choice between "by TEMPLATE" and "by SERIES", select "by SERIES." Then enter the series name (here CCIURR) and specify the starting and ending years you wish to retrieve. Maybe this time, you want 1980 through 1990, so you would enter 80, then enter 90.
EDIT
EXPORT
QUIT
7. You will then be invited to choose a format for your download of a series. If you are going to be reading the data in via SHAZAM, select "PRN format," rather than "TSP format." The file you create will then be called CCIURR.PRN and it will be written to whatever drive you identified at the beginning of the process (here, a: or c:).
8. Now enter the SHAZAM menu and choose TED to take a look at the variable you have downloaded. The first line in the file will be the name of the series you have downloaded. You will want to delete this, or use the option / skiplines=1 on the SHAZAM read statement you use to read in this variable (in order to force SHAZAM not to start reading actual data until the second line).
9. Note that if you have four different files downloaded from CITIBASE, each with a single variable in it (suppose each is monthly data for the same ten-year period), you can read them in successively by using:
sample 1 120You could then use the STAT command to verify the contents of the computer's memory at this point. It would give you the descriptive statistics of each variable and the number of observations. Then you could start doing regressions.
read(cciurr.prn) cciurr / skiplines=1
read(var2.prn) var2 / skiplines=1
read(var3.prn) var3 / skiplines=1
read(var4.prn) var4 / skiplines=1
| COURSE OUTLINE | LECTURE OUTLINES | PROBLEM SETS | PROBLEM SOLUTIONS | COMPUTER LABS |
| SHAZAM EXAMPLES | DATA SETS | ONLINE QUIZZES | GRAPHICS | HANDOUTS |