two graphs

[Olivia Emma]

Hello


I want to combine these two graphs into one graphs please

twoway (scatter ed yearofbirth) || (lfit ed yearofbirth if yearofbirth>=1967, lstyle(dash)) || (lfit ed yearofbirth if yearofbirth < 1964, lstyle(dash)) if rural==1 , xline(1967 1965) legend(off) title(Years of School) xtitle(Birth year )
graph save ed_rural, replace




twoway (scatter ed yearofbirth) || (lfit ed yearofbirth if yearofbirth>=1967, lstyle(dash)) || (lfit ed yearofbirth if yearofbirth < 1964, lstyle(dash)) if rural==0 , xline(1967 1965) legend(off) title(Years of School) xtitle(Birth year )
graph save ed_urban, replace



data



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

* Example generated by -dataex-. For more info, type help dataex
clear
input int yearofbirth float(rural edu)
1945 0  5.461538
1945 1 3.0535715
1946 0 4.4444447
1946 1  2.981132
1947 0 4.4583335
1947 1 3.0229886
1948 0  7.466667
1948 1  3.486111
1949 0  6.466667
1949 1 3.4285715
1950 0  5.333333
1950 1  3.837838
1951 0  7.793103
1951 1  3.505618
1952 0      7.36
1952 1  3.639535
1953 0  6.193548
1953 1         4
1954 0  6.254546
1954 1  3.862857
1955 0      6.86
1955 1  3.469388
1956 0  7.218391
1956 1 3.8533835
1957 0  6.893617
1957 1 4.0705395
1958 0  6.971428
1958 1 4.1626296
1959 0  7.011765
1959 1  3.993243
1960 0    7.4375
1960 1 3.8481014
1961 0  7.695652
1961 1 4.3138685
1962 0  7.971292
1962 1 4.3483872
1963 0  8.376344
1963 1  4.818792
1967 0  9.774648
1967 1  7.380208
1968 0  9.903571
1968 1  7.065299
1969 0  9.743494
1969 1  7.844485
1970 0  9.916924
1970 1  7.581955
1971 0 10.474684
1971 1  7.606625
1972 0 10.030556
1972 1  7.788915
1973 0 10.155216
1973 1  7.918845
1974 0  9.981396
1974 1  7.819808
1975 0   10.1309
1975 1  7.725707
1976 0 10.116773
1976 1  7.886999
1977 0   10.1833
1977 1  7.752955
1978 0  10.25701
1978 1   7.91217
1979 0 10.060362
1979 1  7.930493
1980 0 10.401316
1980 1  8.386111
1981 0 10.675324
1981 1  8.188822
1982 0 10.276224
1982 1  8.335929
1983 0 10.234568
1983 1   8.07248
1984 0 10.370723
1984 1  8.129973
1985 0  10.37156
1985 1  8.451258
1986 0     10.67
1986 1  8.335877
1987 0  10.58913
1987 1  8.426045
1988 0 10.391775
1988 1  8.287082
1989 0 10.039095
1989 1   8.25811
1990 0 10.204603
1990 1  8.161189
1991 0 10.863905
1991 1  8.934037
1992 0  10.68661
1992 1 8.7576475
1993 0  10.43131
1993 1  8.633952
1994 0 10.023728
1994 1  8.825364
1995 0  9.856209
1995 1  8.348605
1996 0 10.740332
1996 1  9.220834
1997 0  10.17801
1997 1   8.86036
end

[Carlo Lazzaro]

Olivia:
you may want to try:

. twoway (scatter ed yearofbirth) || (lfit ed yearofbirth if yearofbirth>=1967, lstyle(dash)) || (lfit ed yearofbirth if yearofbirth < 1964, lstyle(dash)) if rural==1 , xline(1
> 967 1965) legend(off) title(Years of School) xtitle(Birth year ) saving(1)

file 1.gph saved

. twoway (scatter ed yearofbirth) || (lfit ed yearofbirth if yearofbirth>=1967, lstyle(dash)) || (lfit ed yearofbirth if yearofbirth < 1964, lstyle(dash)) if rural==0 , xline(1
> 967 1965) legend(off) title(Years of School) xtitle(Birth year ) saving(2)

file 2.gph saved

. gr combine 1.gph 2.gph

.

[Nick Cox]

My guess differs from that of Carlo Lazzaro. I find the parentheses syntax for different twoway calls much harder to read and write than the alternative. I think you need a legend or at least some other way to distinguish rural and urban.

* Example generated by -dataex-. For more info, type help dataex
clear
input int yearofbirth float(rural edu)
1945 0  5.461538
1945 1 3.0535715
1946 0 4.4444447
1946 1  2.981132
1947 0 4.4583335
1947 1 3.0229886
1948 0  7.466667
1948 1  3.486111
1949 0  6.466667
1949 1 3.4285715
1950 0  5.333333
1950 1  3.837838
1951 0  7.793103
1951 1  3.505618
1952 0      7.36
1952 1  3.639535
1953 0  6.193548
1953 1         4
1954 0  6.254546
1954 1  3.862857
1955 0      6.86
1955 1  3.469388
1956 0  7.218391
1956 1 3.8533835
1957 0  6.893617
1957 1 4.0705395
1958 0  6.971428
1958 1 4.1626296
1959 0  7.011765
1959 1  3.993243
1960 0    7.4375
1960 1 3.8481014
1961 0  7.695652
1961 1 4.3138685
1962 0  7.971292
1962 1 4.3483872
1963 0  8.376344
1963 1  4.818792
1967 0  9.774648
1967 1  7.380208
1968 0  9.903571
1968 1  7.065299
1969 0  9.743494
1969 1  7.844485
1970 0  9.916924
1970 1  7.581955
1971 0 10.474684
1971 1  7.606625
1972 0 10.030556
1972 1  7.788915
1973 0 10.155216
1973 1  7.918845
1974 0  9.981396
1974 1  7.819808
1975 0   10.1309
1975 1  7.725707
1976 0 10.116773
1976 1  7.886999
1977 0   10.1833
1977 1  7.752955
1978 0  10.25701
1978 1   7.91217
1979 0 10.060362
1979 1  7.930493
1980 0 10.401316
1980 1  8.386111
1981 0 10.675324
1981 1  8.188822
1982 0 10.276224
1982 1  8.335929
1983 0 10.234568
1983 1   8.07248
1984 0 10.370723
1984 1  8.129973
1985 0  10.37156
1985 1  8.451258
1986 0     10.67
1986 1  8.335877
1987 0  10.58913
1987 1  8.426045
1988 0 10.391775
1988 1  8.287082
1989 0 10.039095
1989 1   8.25811
1990 0 10.204603
1990 1  8.161189
1991 0 10.863905
1991 1  8.934037
1992 0  10.68661
1992 1 8.7576475
1993 0  10.43131
1993 1  8.633952
1994 0 10.023728
1994 1  8.825364
1995 0  9.856209
1995 1  8.348605
1996 0 10.740332
1996 1  9.220834
1997 0  10.17801
1997 1   8.86036
end


local opts xline(1967 1965) title(Years of School) xtitle(Birth year) xla(1945(5)1995) legend(order(4 "urban" 1 "rural" ) pos(11) ring(0))

twoway scatter ed yearofbirth if rural==1, ms(Oh) mc(blue) /// 
|| lfit ed yearofbirth if yearofbirth>=1967 & rural==1, lstyle(dash) lc(blue) ///
|| lfit ed yearofbirth if yearofbirth<1964 & rural==1, lstyle(dash) lc(blue)  ///
|| scatter ed yearofbirth if rural==0, ms(+) mc(red)                          ///
|| lfit ed yearofbirth if yearofbirth>=1967 & rural==0, lstyle(dash) lc(red) ///
|| lfit ed yearofbirth if yearofbirth < 1964 & rural==0, lstyle(dash) lc(red) `opts'

[Olivia Emma]

Thanks @Nick Cox this is what I am looking please

is there a way to estimate the jump please in each case(rural , urban) please

[Nick Cox]

Many ways. Some large fraction of applied work in economics is currently focused on such jumps, but I am not qualified to suggest which methods might work well or even best in your set-up. I'd ask a new question if you don't get a good answer.

[Olivia Emma]

I mean we can see that in case of urban on average the jump is approximately 2. how we can show this please in graph or calculate this please for the other case

[Nick Cox]

I am not following you. The jump looks very clear to me. How else would it be shown? Adding extra lines or text would just complicate the graph. You might be better off calculating the gaps and commenting in a caption to the graph.

[Olivia Emma]

Nick Cox
local opts xline(1967 1965) title(Years of School) xtitle(Birth year) xla(1945(5)1995) legend(order(4 "urban" 1 "rural" ) pos(11) ring(0)) can you please let me know what does the 4 "urban" 1 "rural" means please , because in my data rural is 1 and urban is 0. Thanks

[Nick Cox]

Start with the help!

help legend option

shows that in my order() call 4 refers to the 4th variable plotted and 1 to the 1st variable plotted. The 4th variable plotted is ed and 1st variable plotted is also ed but the 4th variable is plotted only for urban observations and the 1st variable for only for rural. The legend specification has nothing to do with the values of the variables. That explains the text in the legend. I show urban above rural in the legend, because the urban values are higher.

[Olivia Emma]

@Nick Cox
I am trying the same code for the overall data but not getting the same result please can you please help



local opts xline(1967 1965) title(Years of School) xtitle(Birth year) xla(1945(5)1995) legend() pos(11) ring(0))

twoway scatter ed yearofbirth , ms(Oh) mc(blue) ///
|| lfit ed yearofbirth if yearofbirth>=1967 , lstyle(dash) lc(blue) ///
|| lfit ed yearofbirth if yearofbirth<1964 , lstyle(dash) lc(blue) ///
|| scatter ed yearofbirth , ms(*) mc(red) ///
|| lfit ed yearofbirth if yearofbirth>=1967 , lstyle(dash) lc(red) ///
|| lfit ed yearofbirth if yearofbirth < 1964 , lstyle(dash) lc(red) `opts'

graph save ed, replace



Data


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

* Example generated by -dataex-. For more info, type help dataex
clear
input int yearofbirth float(edu literacy)
1945 3.5072465  .6376812
1946 3.1935484  .5967742
1947  3.333333  .5585586
1948  4.172414  .6666667
1949 4.2280703  .6403509
1950 4.1037035  .3407407
1951  4.559322  .3813559
1952 4.4774776  .4684685
1953  4.626728   .437788
1954 4.4347825  .4086956
1955 4.3299494  .4111675
1956 4.6827197  .4929178
1957  4.862687   .519403
1958  4.911168 .52284265
1959 4.6666665  .5249344
1960  4.787383  .5233645
1961  5.446602   .526699
1962   5.47181 .56973296
1963  5.864139  .5908373
1964  5.849765  .5978091
1965  6.780075  .6484963
1966    7.5573  .6923077
1967  8.398204  .7410179
1968  8.039216  .7230392
1969  8.465937  .7396594
1970  8.348485  .7444444
1971  8.740926  .7546934
1972  8.456954  .7574503
1973  8.703571  .7455357
1974  8.623163  .7675022
1975  8.602033  .7818608
1976  8.698609  .7836167
1977  8.645475  .7584144
1978  8.731428  .7665306
1979  8.692585  .7638589
1980  9.167517  .7253401
1981  9.210855  .7224199
1982  9.090415  .7457512
1983  8.917931       .72
1984  9.050781  .8015625
1985  9.232276  .9580224
1986   9.34632   .965368
1987  9.345656  .9630314
1988  9.179981  .9540863
1989  8.982427  .9615062
1990  8.936893  .9495146
1991  9.843794  .9595537
1992  9.630155  .9639176
1993  9.449275   .957971
1994  9.280928  .9626288
1995  8.919555  .9418317
end

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

[Olivia Emma]

Please see the attached file I donot want to have legend in this please

[Nick Cox]

The help is your friend. pos() and ring() are suboptions of legend() and must appear inside the legend() call.

legend() pos(11) ring(0))

is quite illegal.

legend(pos(11) ring(0))

would be legal.

That said, I don't understand what you're trying to do, as each data point and each fitted line now appears twice.

If you don't want a legend, then don't ask for one in the code or switch it off with legend(off).

Please don't post .gph attachments. This is explained in the FAQ Advice.

[Nick Cox]

Posts are crossing, which is OK. Please see revisions to #12.

What's not so OK: Please see https://www.statalist.org/forums/help#before and the whole of that document. You are expected to try looking at the help!

[Olivia Emma]

Thanks Nick Cox

I will try to do search help before posting

Yes you are right each data point and each fitted line now appears twice. By mistake I did that Please