Correlation output with two decimals

[Alejandro Torres]

Dear statalisters,

When I run the code

pwcorr varlist, star (0.05) .

the output are with 4 decimals.

My questions are, are there any code to get the output with two decimals only? (rounded to the second decimal) and,
are there any way to get stars for (0.1 0.05 and 0.01) at the same output?

Thank you so much for your help.

Best regards,

Alejandro

[Bruce Weaver]

Hello Alejandro. How about this?

sysuse auto, clear
quietly pwcorr trunk weight length turn, star (0.05)
matrix list r(C), format(%4.2f)

Output:

. matrix list r(C), format(%4.2f)

symmetric r(C)[4,4]
         trunk  weight  length    turn
 trunk    1.00
weight    0.67    1.00
length    0.73    0.95    1.00
  turn    0.60    0.86    0.86    1.00

[Alejandro Torres]

Dear Bruce,
Thank you very much for your answer, I am going to try it right now and will let you know.
Thank you again,

Alejandro

[Alejandro Torres]

Dear Bruce,
there are not stars in the output, I tried to add star(0.05) after format(%4.2f) but didn't work.
Do you know what could be happening?
Thank you again for your time,

Alejandro

[Bruce Weaver]

You won't be able to show the stars with that method. You might find something useful in this old thread:

• https://www.statalist.org/forums/for...decimal-places

For more info about the -cpcorr- command mentioned there:

net describe cpcorr, from(http://fmwww.bc.edu/RePEc/bocode/c)

HTH.

[Alejandro Torres]

Dear Bruce,

Thank you very much for your help again.
Best regards !!
Paulo

[John Mullahy]

Months ago I posted in the "Wishlist for Stata 16" thread a suggestion that the -cformat- specification used to format regression output should be expanded to apply to other types of results like correlation matrixes. Evidently the suggestion wasn't implemented in v16.0.

[Clyde Schechter]

The American Statistical Association has recommended that the use of the concept of statistical significance be abandoned. See https://www.tandfonline.com/doi/full...5.2019.1583913. The use of statistical significance stars is a particularly egregious use of this now banished concept.

[Bruce Weaver]

Months ago I posted in the "Wishlist for Stata 16" thread a suggestion that the -cformat- specification used to format regression output should be expanded to apply to other types of results like correlation matrixes. Evidently the suggestion wasn't implemented in v16.0.

In fact, that is the first thing I tried when I read the original post. Unfortunately, it did not work.

[Claire Chuter]

Thank you Bruce Weaver for this answer! It really helped me out just now.