How can i save the output of pwcorr and correlate into dta?Thanks a lot!

[Jacky Wei]

I want to compute the correlation of x between different stocks in each half year. There are 20+ half years and 2000+ stocks. (Table 1). Therefore, I reshape the data firstly yh by yh(Table 2) ，and compute the correlation of x (Table 3).

I want to ask two questions:
1.Is there any concise solutions?
2.How can i save the output of pwcorr and correlate into dta?
Thanks for your reply!

Table 1

stkcd	tradedate	yh	x
1	16806	92	0.078212
1	16807	92	0.128066
1	16811	92	0.597885
1	16812	92	0.012668
1	16813	92	0.2803
...	...	...	...
2027	16806	92	0.507343
2027	16807	92	0.299147
2027	16811	92	0.369548
2027	16812	92	0.216952
2027	16813	92	0.525851
	...	...	...
1	17147	93	0.143762
1	17148	93	0.694211
1	17149	93	0.529129
1	17150	93	0.516332
1	17156	93	0.042779
...	...	...	...
2027	17147	93	0.82723
2027	17148	93	0.973761
2027	17149	93	0.122705
2027	17150	93	0.184536
2027	17156	93	0.250978
	...	...	...
1	17280	94	0.354138
1	17281	94	0.918449
1	17282	94	0.315837
1	17295	94	0.112138
1	17296	94	0.47002
...	...	...	...
2027	17280	94	0.731127
2027	17281	94	0.637081
2027	17282	94	0.404287
2027	17295	94	0.923884
2027	17296	94	0.53458
	...	...	...

Table 2

tradedate	x1	x2	...	x2027	yh
16806	0.078212	...	...	0.507343	92
16807	0.128066	...	...	0.299147	92
16811	0.597885	...	...	0.369548	92
16812	0.012668	...	...	0.216952	92
16813	0.2803	...	...	0.525851	92
...	...	...	...	...	...

Table 3

	x1	x2	...	x2027
x1	1	...	...
x2	0.152024	...	...
...	...	...	...	...
x2027	0.001852	...	...	1

[Joro Kolev]

I am not sure what you are going to do with so many numbers, and I am pretty sure that saving correlation matrices as dta files does not make much sense. But doing it is mechanically easy:

. sysuse auto, clear
(1978 Automobile Data)

. correlate price-weight
(obs=69)

             |    price      mpg    rep78 headroom    trunk   weight
-------------+------------------------------------------------------
       price |   1.0000
         mpg |  -0.4559   1.0000
       rep78 |   0.0066   0.4023   1.0000
    headroom |   0.1112  -0.3996  -0.1480   1.0000
       trunk |   0.3232  -0.5798  -0.1572   0.6608   1.0000
      weight |   0.5478  -0.8055  -0.4003   0.4795   0.6691   1.0000


. mat C = r(C)

. clear

. svmat C
number of observations will be reset to 6
Press any key to continue, or Break to abort
number of observations (_N) was 0, now 6

. list, sep(0)

     +-----------------------------------------------------------------------+
     |        C1          C2          C3          C4          C5          C6 |
     |-----------------------------------------------------------------------|
  1. |         1   -.4559489    .0065533    .1112435    .3232065    .5478396 |
  2. | -.4559489           1    .4023404   -.3995853   -.5798199   -.8055198 |
  3. |  .0065533    .4023404           1   -.1479982   -.1572433   -.4003441 |
  4. |  .1112435   -.3995853   -.1479982           1    .6607842    .4794664 |
  5. |  .3232065   -.5798199   -.1572433    .6607842           1     .669138 |
  6. |  .5478396   -.8055198   -.4003441    .4794664     .669138           1 |
     +-----------------------------------------------------------------------+

.

[Jacky Wei]

I am not sure what you are going to do with so many numbers, and I am pretty sure that saving correlation matrices as dta files does not make much sense. But doing it is mechanically easy:

. sysuse auto, clear
(1978 Automobile Data)

. correlate price-weight
(obs=69)

| price mpg rep78 headroom trunk weight
-------------+------------------------------------------------------
price | 1.0000
mpg | -0.4559 1.0000
rep78 | 0.0066 0.4023 1.0000
headroom | 0.1112 -0.3996 -0.1480 1.0000
trunk | 0.3232 -0.5798 -0.1572 0.6608 1.0000
weight | 0.5478 -0.8055 -0.4003 0.4795 0.6691 1.0000


. mat C = r(C)

. clear

. svmat C
number of observations will be reset to 6
Press any key to continue, or Break to abort
number of observations (_N) was 0, now 6

. list, sep(0)

+-----------------------------------------------------------------------+
| C1 C2 C3 C4 C5 C6 |
|-----------------------------------------------------------------------|
1. | 1 -.4559489 .0065533 .1112435 .3232065 .5478396 |
2. | -.4559489 1 .4023404 -.3995853 -.5798199 -.8055198 |
3. | .0065533 .4023404 1 -.1479982 -.1572433 -.4003441 |
4. | .1112435 -.3995853 -.1479982 1 .6607842 .4794664 |
5. | .3232065 -.5798199 -.1572433 .6607842 1 .669138 |
6. | .5478396 -.8055198 -.4003441 .4794664 .669138 1 |
+-----------------------------------------------------------------------+

.

Thank for your answer. Because one correlation matrix to one half year, there are 20+ correlation matrixs in all. These data need to be saved to a dta file. Your answer is the solution to one correlation matrix, not the solution to me. Thanks again.

[Nick Cox]

There is a corr() function in egenmore fromSSC that may help here.

[Jacky Wei]

There is a corr() function in egenmore fromSSC that may help here.

Thank you very much.If the data can be saved as the dta file, it's so much better.