create dataset/frame from previous collect result

[Luis Pecht]

hi,

Is there any way to create a new frame from the previous collect result:

. clear

. sysuse auto
(1978 automobile data)

. collect clear

. collect: qui reg price mpg rep78 headroom weight length turn displacement gear_ratio

. collect layout (colname) (result[_r_b]) (), name(default)

Collection: default
      Rows: colname
   Columns: result[_r_b]
   Table 1: 9 x 1

------------------------------------
                       | Coefficient
-----------------------+------------
Mileage (mpg)          |   -117.2946
Repair record 1978     |    742.5576
Headroom (in.)         |   -594.5623
Weight (lbs.)          |    3.935387
Length (in.)           |   -74.93533
Turn circle (ft.)      |   -208.5507
Displacement (cu. in.) |    16.98242
Gear ratio             |    1645.675
Intercept              |    10076.63
------------------------------------

desired new frame:

. list

     +---------------------------------+
     |                  var   coeffi~t |
     |---------------------------------|
  1. |         Mileage(mpg)   -117.295 |
  2. |     Repairrecord1978    742.558 |
  3. |        Headroom(in.)   -594.562 |
  4. |         Weight(lbs.)    3.93539 |
  5. |          Length(in.)   -74.9353 |
     |---------------------------------|
  6. |      Turncircle(ft.)   -208.551 |
  7. | Displacement(cu.in.)    16.9824 |
  8. |            Gearratio    1645.68 |
  9. |            Intercept    10076.6 |
     +---------------------------------+

thanks,

[Jeff Pitblado (StataCorp)]

There is no collect command that will produce a Stata dataset or frame from a given layout. However, with the above example, you can export to Markdown, then use import delimited to convert the exported table into a Stata frame.

Here is some code that does this.

sysuse auto

collect: qui reg price mpg rep78 headroom weight length turn displacement gear_ratio

collect layout (colname) (result[_r_b]) (), name(default)

* hide the column title so the Markdown only contains the row headers
* and coefficient values
collect style header result, level(hide)
collect preview

collect export table.markdown, replace
* verify the columns are simply row headers and coefficient values
type table.markdown

frame create results
frame results {
    import delimited junk1 xvar coef junk2 using table.markdown, delimiter("|")
    drop junk?
    describe
    list
}

Here is the output from describe and list.

.         describe

Contains data
 Observations:             9
    Variables:             2
------------------------------------------------------------------------------------------------
Variable      Storage   Display    Value
    name         type    format    label      Variable label
------------------------------------------------------------------------------------------------
coef            str24   %24s
result          float   %9.0g
------------------------------------------------------------------------------------------------
Sorted by:
     Note: Dataset has changed since last saved.
.         list

     +--------------------------------------+
     |                     coef      result |
     |--------------------------------------|
  1. |  Mileage (mpg)             -117.2946 |
  2. |  Repair record 1978         742.5576 |
  3. |  Headroom (in.)            -594.5623 |
  4. |  Weight (lbs.)              3.935387 |
  5. |  Length (in.)              -74.93533 |
     |--------------------------------------|
  6. |  Turn circle (ft.)         -208.5507 |
  7. |  Displacement (cu. in.)     16.98242 |
  8. |  Gear ratio                 1645.675 |
  9. |  Intercept                  10076.63 |
     +--------------------------------------+