Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Re: Polychoric PCA error message

From   Nick Cox <>
Subject   Re: st: Re: Polychoric PCA error message
Date   Mon, 25 Feb 2013 01:46:10 +0000

My guess is that Jay meant Pearson correlations as calculated by -correlate-.


Yashin Lin

> Thanks for the response. There are no longer zero cells--I was simply
> not certain of how to double check that I can now use the 1st
> principal component as asset index, given error messages.
> Would it be possible to clarify what was meant by non polychoric correlations?


> I would recommend looking at the original correlations (non
> polychoric) and one and two-way marginal tables. If there is an excess
> of zero cells or other pathological patterns, polychoric won't
> converge. Sometimes you can fix this up by judicious assumptions such
> as adding one to all cells in the table, but this needs to be handled
> with serious care. Search the archives, this was discussed recently.

Yashin Lin

>> While executing polychoric PCA to produce an asset/wealth index, three
>> iterations of the following message appeared:
>>   numerical derivatives are approximate
>>   nearby values are missing
>> I understand that the first principal component should be the wealth
>> index; it contains negative values, with proportion of explained
>> variance = ~25%.
>> Principal component analysis
>>  k  |  Eigenvalues  |  Proportion explained  |  Cum. explained
>> ----+---------------+------------------------+------------------
>>   1 |    7.955020   |    0.248594            |   0.248594
>> Question: Are there any methods to double check that the procedure
>> performed reasonably well and that I can use this 1st PC as an asset
>> index? 25% explained variance and persistence of error messages
>> suggest that I should be cautious.
>> In case it helps, I have copied below a previous issue I have had with
>> the same dataset,\ which led me to remove five variables, resulting in
>> no zeros in correlation matrix.
*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index