04302a3

# 04302a3 - and Principal Axis 2 NOTE In R using cmdscale...

This preview shows page 1. Sign up to view the full content.

302 Assignment 3 2004 HAND IN THE R CODE YOU USED, NEXT TO EACH ANALYSIS, WE CAN THEN TELL WHERE YOU WENT WRONG! LABEL EACH ANALYSIS WITH THE ASSOCIATED QUESTION NUMBER, LABEL ALL GRAPHS APPROPRIATELY AND KEEP GRAPHS TO A READABLE SIZE. Use the fish catch data you used in the last Assignment. Use the data corrected for number of hooks (i.e. tons of fish per hundred thousand hooks), and remember we need to analyse the standardised data. 1) Do a Principal Coordinates on a Manhattan distance matrix calculated from the standardised data. Present: a) a scree diagram, b) a reduced space plot (with years identified) c) A goodness of fit statistic – what percentage of variance is described by the reduced space plot. d) Identify which variables are highly associated with Principal Axis 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: and Principal Axis 2. NOTE: In R using cmdscale() Mardia’s a 1 goodness of fit statistic is hidden in \$GOF. It is the first of the two numbers there. 2) Now do a nonmetric multidimensional scaling on the same distance matrix as in question 1. a) Present a STRESS diagram. How many dimensions would you choose if you were allowed more than two (i.e. not in an assignment)? b) Present a 2 dimensional solution reduced space plot with site identified. What is the STRESS for this 2 D solution? How good is the reduced space plot? c) Compare the reduced space plots from question 1 and question 2 and comment on the similarities between these two reduced space plots and that from a PCA on the same data. d) Identify which variables are associated with the Principal Axes you chose....
View Full Document

## This note was uploaded on 09/24/2009 for the course STATS 731 taught by Professor Renate during the Spring '09 term at Auckland.

Ask a homework question - tutors are online