how to custom a correlation table

[May Blake]

Hi,

I have run a pairwise correlation below and I would like to make it exported to look like the image below. I do not know how to custom this table, any help will be appreciated.


----------------------- copy starting from the next line -----------------------

pwcorr var1 var2 var3 var4 var5 var6 var7 var8 var8 var9 var10 var11 var12 var13 var14 var15 var16 var17 var18, star(.05) bonferroni

------------------ copy up to and including the previous line ------------------

[Andrew Musau]

Correlation matrices are often represented by their upper or lower triangular portions because they are symmetric - and thus you want to avoid redundancy. Here is a way using estout from SSC. Add the option -nostar- to get rid of the stars, if that is what you want.

sysuse auto, clear
pwcorr mpg weight turn displacement headroom length

*REPLICATING PWCORR OUTPUT AND EXPORTING USING ESTTAB
quietly estpost corr mpg weight turn displacement headroom length, matrix

local eqlabels
forval var= 1/6{
    local eqlabels "`eqlabels' (`var')"
}

esttab ., replace b(3) unstack nonum nomtitle not noobs ///
label compress eqlabels(`eqlabels', lhs("Variables"))

Res.:

. pwcorr mpg weight turn displacement headroom length

             |      mpg   weight     turn displa~t headroom   length
-------------+------------------------------------------------------
         mpg |   1.0000 
      weight |  -0.8072   1.0000 
        turn |  -0.7192   0.8574   1.0000 
displacement |  -0.7056   0.8949   0.7768   1.0000 
    headroom |  -0.4138   0.4835   0.4245   0.4745   1.0000 
      length |  -0.7958   0.9460   0.8643   0.8351   0.5163   1.0000 

. 
. *REPLICATING PWCORR OUTPUT AND EXPORTING USING ESTTAB
. quietly estpost corr mpg weight turn displacement headroom length, matrix

. 
. local eqlabels

. forval var= 1/6{
  2.     local eqlabels "`eqlabels' (`var')"
  3. }

. 
. esttab ., replace b(3) unstack nonum nomtitle not noobs ///
> label compress eqlabels(`eqlabels', lhs("Variables"))

----------------------------------------------------------------------------------------------
Variables              (1)          (2)          (3)          (4)          (5)          (6)   
----------------------------------------------------------------------------------------------
Mileage (mpg)        1.000                                                                    
Weight (lbs.)       -0.807***     1.000                                                       
Turn circ..)        -0.719***     0.857***     1.000                                          
Displacem.. in.)    -0.706***     0.895***     0.777***     1.000                             
Headroom (in.)      -0.414***     0.483***     0.424***     0.474***     1.000                
Length (in.)        -0.796***     0.946***     0.864***     0.835***     0.516***     1.000   
----------------------------------------------------------------------------------------------
* p<0.05, ** p<0.01, *** p<0.001

[Andrew Musau]

Row numbers can additionally be added using a loop over variables. The following assumes that variables are labeled (as in the auto dataset).

sysuse auto, clear
local vars mpg weight turn displacement headroom length
pwcorr `vars'

*REPLICATING PWCORR OUTPUT AND EXPORTING USING ESTTAB
quietly estpost corr `vars', matrix

local eqlabs
local coeflabs
local i 0
foreach var of local vars{
    local++i 
    local eqlabels "`eqlabels' (`i')"
    local coeflabs `coeflabs' `var' "(`i') `:var lab `var''" 
}

esttab ., replace b(3) unstack nonum nomtitle not noobs ///
label compress coeflab(`coeflabs') eqlabels(`eqlabels', lhs("Variables"))

Res.:

. esttab ., replace b(3) unstack nonum nomtitle not noobs ///
> label compress coeflab(`coeflabs') eqlabels(`eqlabels', lhs("Variables"))

----------------------------------------------------------------------------------------------
Variables              (1)          (2)          (3)          (4)          (5)          (6)   
----------------------------------------------------------------------------------------------
(1) Mileage (m~)     1.000                                                                    
(2) Weigh..)        -0.807***     1.000                                                       
(3) Turn ..)        -0.719***     0.857***     1.000                                          
(4) Displ.. in.)    -0.706***     0.895***     0.777***     1.000                             
(5) Headr..)        -0.414***     0.483***     0.424***     0.474***     1.000                
(6) Length (in.)    -0.796***     0.946***     0.864***     0.835***     0.516***     1.000   
----------------------------------------------------------------------------------------------
* p<0.05, ** p<0.01, *** p<0.001