extracting estimates from a twoway tabulation

[Madu Abuchi]

How do I extract estimates from a two way tabulation in stata?

Example data:

clear all
set obs 100
gen id=_n
gen x = runiform()
gen xx=(x>0.5)
gen y = runiform() + 1
gen yy =(y>1.7)
gen z = runiform()
gen zz = 0
replace zz = 1 if (z>0.3 & z<0.7)
replace zz = 2 if z>0.6
tab xx yy, col nokey
tab yy zz, col nokey

tab xx yy, col nokey

        yy
    xx    0    1    Total
                    
    0    38    15    53 
        54.29    50.00    53.00 
                    
    1    32    15    47 
        45.71    50.00    47.00 
                    
    Total    70    30    100 
        100.00    100.00    100.00

. tab yy zz, col nokey

         zz
    yy    0    1    2    Total
                    
    0    19    18    33    70 
        79.17    72.00    64.71    70.00 
                    
    1    5    7    18    30 
        20.83    28.00    35.29    30.00 
                    
    Total    24    25    51    100 
        100.00    100.00    100.00    100.00

I want to plot a twoway line graph or something closer of (xx and yy), and also (yy and zz) and use addplot to plot this two on same scale(or graph)

I noticed after -return list- command that these estimates are not stored. Does any one have information on any trick or user written command that I use to extract this estimates?

Thanks in anticipation.

[Nick Cox]

Consider

1. the matcell() option of tabulate.

2. calculating frequencies directly.

. tab yy zz, col nokey  matcell(freq)

           |          zz
        yy |         0          1 |     Total
-----------+----------------------+----------
         0 |        41         27 |        68
           |     69.49      65.85 |     68.00
-----------+----------------------+----------
         1 |        18         14 |        32
           |     30.51      34.15 |     32.00
-----------+----------------------+----------
     Total |        59         41 |       100
           |    100.00     100.00 |    100.00


. mat li freq

freq[2,2]
    c1  c2
r1  41  27
r2  18  14

. bysort yy zz : gen freq = _N

. egen tag = tag(yy zz)

. l yy zz freq if tag

     +----------------+
     | yy   zz   freq |
     |----------------|
  1. |  0    0     41 |
 42. |  0    1     27 |
 69. |  1    0     18 |
 87. |  1    1     14 |
     +----------------+

3. https://www.statalist.org/forums/for...updated-on-ssc