Download collin command for stata and suitable value for vif ?

[Lenny Tran]

I use command findit collin, and scroll to the last, then click


collin from https://stats.idre.ucla.edu/stat/stata/ado/analysis
collin. Collinearity Diagnostics / Philip B. Ender / Statistical Computing
and Consulting / UCLA Office of Academic Computing / [email protected] /
STATA ado and hlp files in the package / distribution-date: 20101123

but then it appear like this
unknown network protocol
https://stats.idre.ucla.edu/stat/stata/ado/analysis/ either
1) is not a valid URL, or
2) could not be contacted, or
3) is not a Stata download site (has no stata.toc file).


how can i download this command, please help

anyway, i am running the fixed model for data panel, and trying to use collin to know the vif, what should be the value for vif suitable for checking collin ?

[ericmelse]

More can be found here on this subject.
To download, run this code:

net describe collin, from(https://stats.idre.ucla.edu/stat/stata/ado/analysis)

and use:

(click here to install)

All the best.

[Lenny Tran]

More can be found here on this subject.
To download, run this code:

net describe collin, from(https://stats.idre.ucla.edu/stat/stata/ado/analysis)

and use:

(click here to install)

All the best.

when i run the code you give, it still give like this

im currrently using stata12

[Nick Cox]

My guess is that you're using an older version of Stata that doesn't support https. Please note from the FAQ

11. What should I say about the version of Stata I use?

The current version of Stata is 15.1. Please specify if you are using an earlier version; otherwise, the answer to your question may refer to commands or features unavailable to you. Moreover, as bug fixes and new features are issued frequently by StataCorp, make sure that you update your Stata before posting a query, as your problem may already have been solved.

and also

6. Why are real names preferred?

You are asked to post on Statalist using your full real name, including given name(s) and a family name, such as "Ronald Fisher" or "Gertrude M. Cox". Giving full names is one of the ways in which we show respect for others and is a long tradition on Statalist. It is also much easier to get to know people when real names are used.

If you overlook this on first registration, it is easy to fix. Click on “Contact us” located at the bottom right-hand corner of every page.

[Lenny Tran]

My guess is that you're using an older version of Stata that doesn't support https. Please note from the FAQ



and also

sorry for the name
im currently using the stata 12, so there is no way to use coliin in stata 12 , right ?

[Nick Cox]

You can't use your Stata to install it directly using net. But you've (we've) got a URL and there are many ways to get at that file, including browsers. Here's the file collin.ado

*! version 1.6    -11/23/10 add sample size
*! version 1.5.2  - 6/17/09 stupid bug fixed
*! version 1.5.1  - 7/16/04
*! version 1.4    - 1/25/04
*! version 1.1.1  - 1/31/02  
*! version 1.1    - 2/7/01
program define collin, rclass
  version 6.0
  syntax varlist [if] [in] [, corr rinv]
  marksample touse
  local vars `varlist'
  tempname r ri r2 x v xr vr ci cir t sscp det diag1
  mat accum `r'    = `vars' if `touse', noconst deviations
  mat `r' = corr(`r')

 /*
  if "`cdiag'" ~= "" {
    quietly mat accum `sscp' = `vars' if `touse', noconst
    mat `diag1' = diag(vecdiag(`sscp'))
    mat `diag1' = syminv(`diag1')
    mat `sscp' = `diag1'*`sscp'*`diag1'
    matrix symeigen `x' `v' = `sscp'
    local str1 "raw sscp (no intercept)"
  } */
  

  if "`corr'"~="" {
    mat `sscp' = `r'
    matrix symeigen `x' `v' = `r'
    local str1 "deviation sscp (no intercept)"
  }
  
  else {
    quietly mat accum `sscp' = `vars' if `touse'
    mat `diag1' = diag(vecdiag(`sscp'))
    local i = 1
    while `i' <= colsof(`sscp') {
      mat `diag1'[`i',`i'] = sqrt(`diag1'[`i',`i'])
      local i = `i'+1
    }
    mat `diag1' = syminv(`diag1')
    mat `sscp' = `diag1'*`sscp'*`diag1'
    matrix symeigen `x' `v' = `sscp'
    local str1 "scaled raw sscp (w/ intercept)"
  }
  
  mat `ri' = inv(`r')
  mat `r2' = vecdiag(`ri')
  mat `t'  = vecdiag(`ri')
  scalar `det' = det(`r')
  local m  = colsof(`v')
  local m2 = colsof(`r')
  local i = 1
  local var2 : colnames(`sscp')
  matrix colnames `v' = `var2'
  matrix rownames `v' = " "
  matrix `ci' = J(1,`m',0)
  local sum = 0
  local i = 1
  while `i' <= `m2' {
    local sum = `sum' + `t'[1,`i']
    matrix `r2'[1,`i'] = 1 - 1/`r2'[1,`i']
    local i = `i'+1
  }
  local i = 1
  while `i' <= `m' {
    matrix `ci'[1,`i'] = sqrt(`v'[1,1]/`v'[1,`i'])
    local i = `i'+1
  }
  local cn  = `ci'[1,`m']
  local mvif = `sum'/`m2'
  tokenize `var2'
  display
  display in green "  Collinearity Diagnostics"
  display  
  display in green "                        SQRT                   R-"
  display in green "  Variable      VIF     VIF    Tolerance    Squared"
  display in green _dup(52) "-"
  local i = 1
  while `i' <= `m2' {
     display in yellow %10s "``i''" %10.2f `t'[1,`i'] /*
            */  %8.2f (sqrt(`t'[1,`i'])) /*
            */ "  " %8.4f (1/`t'[1,`i'])  /*
            */ "  " %10.4f  `r2'[1,`i']  
                
     local i = `i'+1
  }
  display in green _dup(52) "-"
  display in green %10s "Mean VIF" %10.2f in yellow `mvif'
  display
  display in green "                           Cond"
  display in green "        Eigenval          Index"
  display in green _dup(33) "-"
  local i = 1
  while `i' <= `m' {
     display in yellow  "    "  `i' /*
            */ "   "  %8.4f `v'[1,`i']  /*
            */ " " %15.4f `ci'[1,`i']        
     local i = `i'+1
  }
  
  
  display in green _dup(33) "-"
  display in green " Condition Number" %15.4f in yellow `cn' " " %10.4f `cnr'
  display in green " Eigenvalues & Cond Index computed from " "`str1'"
  display in green " Det(correlation matrix)" %10.4f in yellow `det'
  
        
  if "`rinv'" ~= "" {
    display
    display in green "Inverse of correlation matrix"
    mat lis `ri', noh
  }
 
  return scalar m_vif = `mvif'
  return scalar cn = `cn'
end

and collin.hlp

.-
help for ^collin^
.-

Collinearity Diagnostics
------------------------

    ^collin^ varlist  [^if^ exp] [^in^ range] [, ^corr rinv^]



Description
-----------

^collin^ computes several collinearity diagnostic measures including
       VIF, tolerance, eigenvalues, condition index, and R-squared.  
       It will compute the eigenvalues and condition index on either
       the raw SSCP with an intercept (default) or the deviation SSCP
       without an intercept.

Option
------

^corr^   Eigenvalues and condition index computed from correlation matrix
         without a constant.
        
      By default the eigenvalues and condition index are computed on the
      scaled raw score SSCP matrix with an intercept.
        
^rinv^  display inverse of correlation matrix.

Examples
--------

 . ^collin read write math^
 . ^collin read write math, corr rinv^
 . ^collin read write math if gender==1^
 . ^collin read write math if e(sample)
 


Author
------

      Philip B. Ender
      UCLA Department of Education
      UCLA, Academic Technology Services
      ender@@ucla.edu


Also see
--------

 Manual:  ^[R] regression diagnostics^
On-line:  help for @vif@

Stata wants to see those files in the c folder of whatever

adopath

calls PLUS

I don't know where Phil Ender's vif is. I guess it's long since superseded by an official variant.

[Lenny Tran]

great, thanks alot, but i downloaded the ado file for collin before, and i can use the ado file in stata 12 correctly right, or i have to use the ado file collin with stata 14 ?

[Nick Cox]

The program should work with Stata 6 up. I don't know why you think you need Stata 14.

[Lenny Tran]

The program should work with Stata 6 up. I don't know why you think you need Stata 14.

if my model like this
xtreg y x1 x2 x3, fe
then i should command collin like this, right ?
collin x1 x2 x3


Thanks a lot

[Lenny Tran]

if my model like this
xtreg y x1 x2 x3, fe
then i should command collin like this, right ?
collin x1 x2 x3


Thanks a lot

Am i right ?

[Hendrik Steen]

Dear all,

I am struggling to install the Stata command collin. I always type findit collin and then try to install the installation package. However, Stata completely crashes when I click on the install button. I use Stata 14.0 on a MacBook Air. Does anyone have a clue how to fix this issue?

Looking forward to receiving any suggestions. Thanks a lot.

[Miftah Zikri]

Download
https://stats.idre.ucla.edu/wp-conte.../12/collin.zip

or

Find "collin" and other commands:
https://stats.oarc.ucla.edu/stata/ado/analysis/

this link below for other useful commands:
https://stats.oarc.ucla.edu/stata/ado/

[Alicia Johnson]

Just in case others run into this issue - I got this same error message when I was connected to my work's VPN. After disconnecting from the VPN, I ran

findit collin

and was able to click through the menu to install the package successfully.