• Sample Data (download to D:\drive first)
• TUTORIAL

VISUAL BASIC -- The Beginnings

• BASIC Programming in Visual Basic -- Tutorial Lesson #2

• ASCII CHARACTER SET  (You thought I was kidding)
• Download VB Tutorials/Shareware

• SIMPLER -- Shareware VB Tutorial for Download

Here we revisit the Aquarium membership problem with a more compact and "elegant" solution to the problem. In several parts of our original solution, we needed to re-initialize the MembershipCount by MembershipType ARRAY .. and we also wrote out the compiled array in a two or more places.

In this solution, we employ the powerful feature of all modern programming languages -- The SUBROUTINE.  Subroutines or sub-programs are smaller, separate, mini-programs which do special tasks. Often these tasks are needed in different parts of the MAIN PROGRAM or even in different programs (re-usable code).

Visual Basic will rely entirely on subroutines -- we will only write subroutines, the "main program" is generated by the "visual" part -- the Graphical User Interface we build with the visual tools. Actual "work" is done in subroutines .. the parts that we write.

We need to look more carefully at our VARIABLES and decide which will be "local" to a subroutine and which need to be "global" or SHARED.  SHARED variables (declared in the DIM statement) will have their current values known to all subroutines at the same time. Local variables are used within subroutines and do not affect other parts of the program.

Local variables are sometimes "passed" to another subroutine to be used by that program.  This is similar to the "functions" we have used already such as:
 

ITYPE = INT(MTYPE)   .. where INT( ** ) is the function [which returns the integer value of a real number, either single or double] and MTYPE is the "argument" or variable passed to the function.

Notice that the Main Program is first (in VB, we never see the "main program per se") and all the SUBroutines follow.  We can edit them without changing anything in the Main Program -- VB is based on this principle, you edit your SUBroutines and leave the main program alone.

I think the "logic" of solving the membership types by zipcode problem described last week will be clearer with the use of subroutines -- you can see the program in its logical whole easier as much of the "work" is done in the subroutines.

STUDY the program and see if you understand the ways it works. The new program has been added to the ZIP and ready for download.

'-------------------------------------------------------
DECLARE SUB ADDMEMS (MTEMP AS SINGLE, NTEMP AS INTEGER)
DECLARE SUB INITNEW ()
DECLARE SUB MEMDUMP ()

OPTION BASE 1

DIM SHARED MEMTYP(5)  AS SINGLE   'ARRAY OF MEMBERSHIP TYPE COUNTS BY ZIPCODE
DIM SHARED ZIPCODE AS STRING    'ZIPCODE NUMBER AS A CHARACTER STRING
'---------------------------------------------------------
DIM MTEMP AS SINGLE      'TEMPORARY MEMBERSHIP TYPE
DIM ZTEMP AS STRING      'TEMPORARY ZIPCODE
DIM NTEMP AS INTEGER     'TEMPORARY MEMBERSHIP COUNT BY TYPE
DIM FIRST AS INTEGER     'CONTROL OF OUTPUT VARIABLE

' OPEN DISKFILES

OPEN "ZIP_RAW.TXT" FOR INPUT AS #1
OPEN "Z_TEMP.TXT" FOR OUTPUT AS #2

CLS   'CLEAR THE SCREEN
'-------------------------------------------------
' Initialize ZipCode Value and MemTyp Array Prior to Input
   ZIPCODE = "00000"
   FIRST = 1
   CALL INITNEW

' BEGIN DATA INPUT FROM DISKFILE IN ORIGINAL FORMAT
'      ZIPCODE, MemberType, MemberCount

WHILE NOT EOF(1)

  INPUT #1, ZTEMP, MTEMP, NTEMP
    'PRINT ZTEMP, MTEMP, NTEMP  'DEBUG TEMP PRINT --DELETE LATER

    IF ZTEMP <> ZIPCODE THEN  'NEW ZIP CODE

      IF FIRST THEN 'FIRST TIME THRU THE PROCESS -- SAVE INFO ONLY
        ZIPCODE = ZTEMP
        CALL ADDMEMS(MTEMP, NTEMP)
        FIRST = 0
      ELSE 'NEW ZIPCODE ENCOUNTERED, DUMP OLD ZIP MEMBERSHIP INFO
        CALL MEMDUMP
        ZIPCODE = ZTEMP
        CALL INITNEW
        CALL ADDMEMS(MTEMP, NTEMP)
      END IF

    ELSE ' SAME ZIP AS LAST RECORD, ADD INFOMATION TO MEMBERSHIP ARRAY

       CALL ADDMEMS(MTEMP, NTEMP)

    END IF

     ' INPUT KY$ ' DEBUG KEY INPUT -- DELETE LATER

WEND

'END OF INPUT FILE ENCOUNTERED --

     CALL MEMDUMP ' WRITE LAST ZIPCODE'S INFO TO FILE

CLOSE  'ALL FILES
END
 

SUB ADDMEMS (MTEMP AS SINGLE, NTEMP AS INTEGER)
DIM I, J AS INTEGER

  I = INT(MTEMP)
  MEMTYP(I) = NTEMP

END SUB

SUB INITNEW
  DIM I AS INTEGER
    FOR I = 1 TO 5
       MEMTYP(I) = 0
    NEXT I
END SUB

SUB MEMDUMP

  PRINT ZIPCODE, MEMTYP(1); MEMTYP(2); MEMTYP(3); MEMTYP(4); MEMTYP(5)

 'WRITE #2, ZIPCODE, MEMTYP(1), MEMTYP(2), MEMTYP(3), MEMTYP(4), MEMTYP(5)

END SUB
 
 
 

Problem #3: -- REAL WORK

Here is a real problem we encountered last week as part of the Austria-Exchange project with the Aquarium.  We needed to get counts of Membership Types (collapsed into 5 classes) into a Single Table with ZipCode as the Primary Key in order to make a Pie-Chart Map of Membership Types by ZipCode area.

Woody and Suzanne were able to summarize the membership types by ZipCode, but only to the extent that an ACCESS (database) table was produced that logically looked like this:
 

ZIPCODE | MEMCLASS | SUM

A portion of the "raw data" looks like this:

"90071",1.00,1
"90071",3.00,1
"90075",2.00,1
"90075",3.00,1
"90077",1.00,3
"90077",2.00,5
"90077",3.00,20
"90077",4.00,5
"90077",5.00,6
"90081",4.00,2
"90086",2.00,1
"90086",4.00,1
"90089",4.00,1
"90201",1.00,1
"90201",2.00,3
"90201",3.00,16
"90201",4.00,2
"90202",3.00,1
"90209",4.00,1
"90210",1.00,8
"90210",2.00,2
"90210",3.00,25
"90210",4.00,7
"90210",5.00,3

Getting the data into an text file allowed me to write a short program to read the data and output into the desired format, which logically looks like this:
 

ZIPCODE | TYPE1 | TYPE2 | TYPE3 | TYPE4 | TYPE5

"90071",1,0,1,0,0
"90075",0,1,1,0,0
"90077",3,5,20,5,6
"90081",0,0,0,2,0
"90086",0,1,0,1,0
"90089",0,0,0,1,0
"90201",1,3,16,2,0
"90202",0,0,1,0,0
"90209",0,0,0,1,0
"90210",8,2,25,7,3

The BASIC program contains a number of "tricks" which make the conversion pretty simple -- once you understand the underlying logic. STUDY this program. Not for its elegance, but for an example of how to work through the knarly logic necessary to solve the problem.

Try writing your own version. Viele Gluck.
 

#2

• In this exercise, we'll write a really useful program to format downloaded data  into two files for import into ArcView.
• This involves using INPUT from a file and WRITE to new files in ASCII format.
• To import into ArcView, we need TWO files:
• A "Geography File" which has a Unique ID for each feature (usually points)
• An "Attribute File" which has the Unique ID along with the data variables.
• These files have general formats which must be matched in order to import correctly:
• A "header record" which names the columns in the data base, like
"ID", "ATTR1", "ATTR2", "ATTR3" ...
• and DATA RECORDS which contain the actual ID and attribute values, like
1, 12.5, 333.34, 128.9, ....
• File "I/O" (input/output) is done by OPENing and CLOSEing diskfiles to read the data in its original form and write out the information in the required formats (above).
• Subsequent read (INPUT #n) and write (WRITE #n) statements (where 'n' is the "file number" used by the program to refer to a specific disk file) will perform the heart of the program core.
• Other calculations -- converting units, creating new derived variables, changing data from one type to another (eg. from Character String to Numbers) -- may follow the INPUT and precede the WRITE statements in the program.

Consider a downloaded file (cities.txt) with City locations (lon/lat) and Population Attributes. Create a TEXT file (notepad).
 
"Los Angeles", -118.15, 34.05, 8235942, 42.1
"Long Beach", -118.05, 33.85, 792435, 37.5
"Santa Ana", -117.9, 33.9, 627962, 67.4

 
HERE IS a sample program which reads the original file and creates the two new files for ArcView:
 
DIM CITY AS STRING
DIM LON, LAT AS SINGLE
DIM POP AS DOUBLE
DIM SPKSPAN AS SINGLE

DIM NUMREC AS INTEGER

OPEN "CITIES.TXT" FOR INPUT AS #1
OPEN "CITYXY.TXT" FOR OUTPUT AS #2
OPEN "CITYDAT.TXT" FOR OUTPUT AS #3

NUMREC = 0
CLS

WRITE #2, "ID" , "LON", "LAT"  ' Header Record indicates data column names for Arcview
WRITE #3, "ID", "CITY", "POP", "SPKSPAN"  'Header Record for the Attribute File

WHILE NOT EOF(1)  ' LOOP UNTIL END OF FILE ON INPUT -- NO MORE DATA
 

INPUT #1, CITY, LON, LAT, POP, SPKSPAN
NUMREC = NUMREC + 1

PRINT CITY; LON; LAT; POP; SPKSPAN   'PRINT TO SCREEN
WRITE #2, NUMREC, LON, LAT  ' write X-Y coords to GeoFile
WRITE #3, NUMREC, CITY, POP, SPKSPAN  ' write Attributes to Datafile
 

WEND  'END OF WHILE
' NEXT STATEMENT AFTER END OF DATA LOOP
 
PRINT "NUMBER OF RECORDS PROCESSED = "; NUMREC
CLOSE   'CLOSE ALL OPEN FILES
END

• Debug (if necesary)

• IMPORT your data into ArcView and map it!

• modify the above to prompt the user for the file names.
Hint:
INPUT "Enter the Original File Name "; FNAME1
OPEN FNAME1 FOR INPUT AS #1
• DOWNLOAD your own raw dataset from the web. Figure out the data format, name your data fields and construct your DIM data definitions for the data you have to read.  Modify the program to read and write YOUR data to new files.

• Write a program which
• Prompts the user to enter the RADIUS of a Buffer around a point IN FEET
Hint:  INPUT "Enter the Buffer Radius "; BUFLEN
(check actual syntax of "prompting" the user for keyboard input in the INPUT statement)
 
• Calculates the AREA of the circle of that radius (sqft)
• Prints out the results (on the screen) in the form:
 
PRINT "The AREA of the Buffer around the point is "; AREA; "."
(check actual syntax for formatting literal string output in the PRINT statement)
• Modify the program above to loop five times prompting for lengths and printing out the area
• Clear the screen before the first prompt (CLS)
• Don't CLS clear screen between input/outputs
• Include the Radius in the output line -- eg. "Buffers with a radius of "; BUFLEN; " feet will have an Area of...

• Modify your program to convert the area to Square METERS.  Show both units (feet and meters) in the final output.