search

philchalmers / faoutlier

Detecting outlying and influential cases for factor analysis in R
GNU General Public License v2.0
7 stars 4 forks source link

faoutlier - forward.search, why 10 values? #2

Closed yashbhambhani closed 8 years ago

yashbhambhani commented 8 years ago

So I'm trying to use the faoutlier package and forward.search function to run an outlier analysis in an EFA dataset, and in default mode [forward.search(data, number of factors)], it returns only 10 values - and it returns only 10 values if i specify cri="mah" (mahalanobis) as well. Why is that so? And when its running in default, what is it actually running? Because running default and mahalanobis is giving me different answers

philchalmers commented 8 years ago

Are you certain it's returning the top 10, or just printing those? The default prints the top extreme cases, but you can used other methods to display the rest of the case values.

yashbhambhani commented 8 years ago

Hi Dr. Chalmers - Thank you for your comment. Its simply printing 10 values - attached is a screenshot of the output. I'm also having trouble understanding what it is running in the default option - mahalanobis distance along with something else?

screen shot 2016-04-01 at 9 13 37 pm
philchalmers commented 8 years ago

The GOF statistic as well, but others can also be included. See the ?GOF help file for details. It might be good to work through those functions first before experimenting with a forward.search function (which is really only useful if you think there are 'clusters') of outliers. Cheers.

yashbhambhani commented 8 years ago

Ok - i will. Thank you for the help!

seonghobae commented 8 years ago

Hi, Phil. Can I select outliers these tuples in this an issue? (160, 162, 77, 169, 203, 258, 3, 316, 341, 35) I can't understand even I read a manual.

Seongho