Welcome to the COMP150DW Web Page

Spring 2004

CONTACT || ANNOUNCEMENTS || LECTURES || HOMEWORK || LINKS

Announcements

o Assignments 1-3 grades are now available via 'progress'
o The command psql -h pgsql may not be working. To access pgsql type (from sun or moon): ssh pgsql
When prompted, enter your password, and when you get the command prompt, type: psql
o There is a mailing list for the class. If you have not received messages from it, please email majordomo@cs.tufts.edu with the command 'subscribe comp150dw' in the body of the message.

Contact Information

Teacher: Professor Daniel Hebert - dhebert@mitre.org

TA: Kate Butler - kbutlr@cs.tufts.edu
Office Hours: Monday 3-4pm, Wednesday 4-5pm (Halligan 118)

Homework
HW1 || HW2 || HW3 || PROJECT || GENERAL INFO

Project :: DUE MAY 13
o Informational document and zip file located here
o Submit with 'provide comp150dw project'
o Note: All data must be entered into the tables. However, the program need not use it all.
o Please use 'provide' for as much of your information as possible: This includes .doc or .ppt files. If you turn something into my mailbox it MUST be there on Friday morning when I come to pick it up -- the projects will be graded that weekend and I won't be able to get at the office after Friday

Deliverables:
o Your Data Warehouse schema
o ALL data in your tables
o The program code
o A readme file with instructions on how to compile, run, etc., plus any information you feel might be helpful in addition.

Homework 3 :: DUE APRIL 22 BEFORE CLASS

o This homework assignment will utilize the data warehouse you previously built incorporating the hurricane data
o Implement an automated Attribute-Oriented Induction capability on your hurricane data
o Input needed :

Your relation tables for the hurricane data
A DMQuery for characterization
A list of attributes
A set of concept hierarchies or generalization operators
A generalized relation threshold
Attribute generalization thresholds for each attribute

o Transform the DMQL statement to a relational query (can do this by hand)
o Use this relational query in your program to retrieve data and then perform the AOI (see algorithm on pg. 188 of book)
o Visualize the results of the AOI via a generalized relation

Deliverables:
o The DMQL statement
o The program code
o A readme file with instructions on how to compile, run, etc., plus any information you feel might be helpful in addition.

Hints:
o This assignment requires just one query.
o After performing the AOI the result can be output in a table. You don't need to output the results from the original query -- just the results after the AOI.
o Even if your HW2 tables had some problems, you probably don't need to redo them to finish this assignment. The data won't be regraded.
o Input does not mean literal input to your program. It just means things you need to be aware of before you jump in to writing it.
o Interactivity is not a requirement.

Homework 2 :: DUE MARCH 18 BEFORE CLASS

o Design a data warehouse based on the hurricane data provided (three years of data). Two data sources with slightly different formats and data
o Utilize a star or snowflake schema
o Implement the data warehouse utilizing relational tables in postgres
o Integrate the data from the two separate data sources and import into your data warehouse
o Keep the data warehouse as you will utilize it in your next homework assignment

Hints
o There is no one right answer to this assignment.

Deliverables
o The schema for the data warehouse.
o The tables themselves (in PostGres).

Homework 1 :: DUE MARCH 11 BEFORE CLASS

o Implement a Data Warehouse using a star schema for the sales example in Chapter 2 page 49
o Import the following data into your database. SalesFact.txt, BranchDimension.txt, ItemDimension.txt, LocationDimension.txt, TimeDimension.txt (data files)
o Use “\COPY” command in postgresql to copy data
o Implement a program that allows you to do the following :

roll-up on total sales by year and by year/quarter
roll-up on total sales by country, by state, and by city
roll-up on total sales by item brand and by item type (digital or analog)
drill down on total sales by month and by day
drill down on total sales by item name
drill down on total sales by street address

o Have a simple user interface to allow activation of each of the functions
o Use SQL statements to implement each roll-up, drill-down

Hints
o Example - Run drill down on total sales by street address in Query Analyzer: select sum(s.dollarssold) as totalsales, l.street from SalesFactTable s, LocationDimensionTable l where l.locationkey = s.locationkey group by l.street
o You can run psql from any of the unix machines with: psql -h pgsql
o \?, \h and \h [command] in psql will provide information
o Here's an example of a c program which interfaces with psql: test1.c. To compile it, log into pgsql and use gcc -lpq.
o The syntax of the copy command is: \copy [table name] from '[file name]' using delimiters '[delimiter]'
o If you're getting errors running PostGres from a remote host, try logging in to pgsql directly.

How to turn in your assignments

To turn in your postgres assignments type the following:
provide comp150dw hwX filenames
where filenames are the names of the files you want to submit, and X is the number of the assignment.
Start with X = 1
You must always submit all your code and executables.

How to check your grades on assignments

To check your grades for your assignments, type:
progress comp150dw

Remember
To get credit for your assignments, you need to grant the TA permission to access your tables. This can be done with the following psql command:
'grant all on table to kbutlr;'
where table is a comma seperated list of your tables' names