Dynamic Factor Analysis Procedure

[jhkim7]

Dear SAS Users

The data consists only of x and I want to do a Dynamic Factor Analysis

I would like to know the concept of Dynamic Factor Analysis and the corresponding SAS Procedure.

 - SAS Procedure, Example

Thank you for your help.

jhkim

[PaigeMiller]

PROC SSM

See:

https://communities.sas.com/t5/SAS-Programming/Command-for-Dynamic-Factor-Analysis/m-p/325401#M26093... (and the next message)

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/etsug/etsug_ssm_examples07.htm

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/etsug/etsug_ssm_examples11.htm

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/etsug/etsug_ssm_examples14.htm

[jhkim7]

Please ask further questions.

Looking at the SSM manual you gave me, I think that both Y and X values ​​are required.

What I want to do is apply data consisting of only X values
I want to find the final common factor through DFA.

If you want to do the above, please ask how to proceed.

If it is SSM, at what stage should I apply it? Or is there another SAS procedure?

Thank you for your help.

jhkim

[ballardw]

---

@jhkim7 wrote:

Please ask further questions.

Looking at the SSM manual you gave me, I think that both Y and X values ​​are required.

What I want to do is apply data consisting of only X values
I want to find the final common factor through DFA.

If you want to do the above, please ask how to proceed.

If it is SSM, at what stage should I apply it? Or is there another SAS procedure?

Thank you for your help.

jhkim

---

Maybe you should be providing an example and interpretation of what you think a "factor analysis" of a single variable would look like. Or what you mean by "factor".

There are multiple procedures in SAS that do factor analysis but perhaps you meant something other than formal factor analysis in those terms.

[jhkim7]

---

@ballardw wrote:

---

@jhkim7 wrote:

Please ask further questions.

Looking at the SSM manual you gave me, I think that both Y and X values ​​are required.

What I want to do is apply data consisting of only X values
I want to find the final common factor through DFA.

If you want to do the above, please ask how to proceed.

If it is SSM, at what stage should I apply it? Or is there another SAS procedure?

Thank you for your help.

jhkim

---

Maybe you should be providing an example and interpretation of what you think a "factor analysis" of a single variable would look like. Or what you mean by "factor".

There are multiple procedures in SAS that do factor analysis but perhaps you meant something other than formal factor analysis in those terms.

---

I am in charge of implementing the contents of research in the research institute into SAS

The lab uses Python, and there is a DFM model in Python, and it was analyzed in this way.

The data consists of only the X variable, and the first principal component common factor was calculated using Python DFM.

I was worried about the implementation of this process, and I made an inquiry.

So, is the procedure SSM or Factor?

Thank you for your help.

jhkim

[PaigeMiller · Accepted solution]

The data consists of only the X variable, and the first principal component common factor was calculated using Python DFM.

This sentence makes no sense. If there is only one X variable (you said "variable" and not "variables") then the principal component on one X variable would return the exact same variable X (although it may be centered and scaled).

Further explanation is required before any more advice can be provided.

[jhkim7]

It seems that there was confusion in the process of my explanation. Sorry about this. 

I am in the process of converting the customer's research in Python into SAS.

The lab uses Python, and the DFM(or DFA) model was analyzed in Python.

* DFM (Dynamic Factor Model), DFA(Dynamic Factor Analysis)

Data consists of only X variables, and it is a process of finding the first principal component common factor by applying

 (Ex) Data =>  yy-mm-dd  x1 x2 x3 ~ x30

I would like to ask you if there is a procedure or method to do DFM (or DFA) analysis in SAS if it consists of only X variables. (SSM or Factor Procedure? Option? )

Thank you for your interest and reply,

jhkim

[Reeza]

Show your python code and at least mentioning the package may help a bit.

I think you want proc factor but hard to say. And to get the results to match will definitely need you to verify that the assumptions match as that's not guaranteed even if you use the same 'terms'. By definition, factor analysis doesn't always have a unique solution as well. 

EDIT: see this post, SSM seems to be the right procedure.

https://communities.sas.com/t5/SAS-Programming/Command-for-Dynamic-Factor-Analysis/td-p/134668

https://documentation.sas.com/doc/en/pgmsascdc/v_034/etsug/etsug_ssm_examples07.htm

[ballardw]

---

@jhkim7 wrote:

It seems that there was confusion in the process of my explanation. Sorry about this. 

I am in the process of converting the customer's research in Python into SAS.

The lab uses Python, and the DFM(or DFA) model was analyzed in Python.

* DFM (Dynamic Factor Model), DFA(Dynamic Factor Analysis)

Data consists of only X variables, and it is a process of finding the first principal component common factor by applying

 (Ex) Data =>  yy-mm-dd  x1 x2 x3 ~ x30

I would like to ask you if there is a procedure or method to do DFM (or DFA) analysis in SAS if it consists of only X variables. (SSM or Factor Procedure? Option? )

Thank you for your interest and reply,

jhkim

---

Compare the text highlighted above in blue with this quote from your original statement:

The data consists only of x 

and you should see where the confusion started. The first statement only mentions exactly one variable.

I don't speak Python and with something like " x1 x2 x3 ~ x30" I would be pretty sure there are at least 4 variables and possibly 30 whose names start with X.