Exercise 2b: Matrix Multiplication

Post date: Jan 26, 2014 4:26:55 PM

For general instructions on completing the exercises, visit the Exercises page.

In this exercise, we will be using matrix multiplication. The best way to do that in UCINET is to use the Command Line interface, also known as the Matrix Algebra facility. On the main menu, choose the 'ALG' button (or press Ctrl-G). Once you are there, you can multiply two matrices A and B by typing

->C = prod(A, B)

->dsp C

Commas or spaces or both can be used to separate the A dataset from the B. If the dataset names contain spaces then you need to enclose them in full quotes, as in:

->MyProduct = prod("first dataset", "second dataset")

You will also need to transpose a matrix. In matrix algebra you can do this like this:

->B = Transp(A)

That means, take the matrix in A, transpose the rows and columns, and save the result as a new dataset B. You can also do it from the UCINET main menu by going to Data|Transpose

We will be using the PADGETT dataset. This dataset contains two matrices, PADGM and PADGB. To use them in matrix algebra, we will first need to separate them into different datasets. You can do this in matrix algebra by typing

->padgm = lev(padgett 1)

->padgb = lev(padgett 2)

You can also do this instead:

->unpack padgett

which will automatically create padgm and padgb.

Or, in the UCINET menu, go to do Data|Unpack. Enter Padgett as the input dataset. Then make sure that "prefix for output files" is set to blank. The press Ok. Both of these procedures will create as output two datasets, called padgm and padgb respectively. Padgm represents marriages ties and padgb represents business ties.

1. Multiply padgm by padgb (i.e., prod(padgm, padgb)). Call the result mb.

2. Display the contents of mb. If there is a zero in the (2,6) cell (Albizzi to Ginori), what does that mean? Interpret the 1s, 2s and 3s as well.

3. Multiply padgb by padgm. Call the results bm. Yes, order matters. Think of mb and bm as two newly measured social relations among these families. What would you call these relations? What does it mean for a family to have the mb relation with another family? How is it different from having the bm relation with the other family?

4. Multiply padgm by padgm and call the result mm. Display mm. How do you interpret the values?

5. Switching datasets, multiply the dataset campnet by its transpose. How do you interpret the values?