I am trying to acquire a bivariate density contour plot of fathers' and their sons' income. I want to display contour lines which represent innermost 95, 90,75, 50, 25 percent of the marginal distributions. My data is shown below:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input int cage float(chour_wage cincom_a) int age float(hour_wage incom_a) 6 . 7.212492 11 . 8.800091 4 . . 10 . 7.767715 4 . 7.096818 8 4.247268 8.301566 4 . 6.221349 8 4.662684 8.35622 6 5.064219 8.718682 11 . 8.904714 5 . . 11 . 9.126666 5 . 8.295078 11 . 8.868837 6 5.506705 8.652475 10 . 8.225165 6 . . 13 . 8.651724 4 . . 9 4.5149555 8.82954 5 . . 12 . 8.760176 4 . . 11 4.6266828 9.257171 4 . . 9 5.899766 9.591916 4 . . 10 . 8.358793 4 2.1330256 5.939688 10 . 8.462613 5 . . 11 . 8.740336 6 4.769462 8.620869 12 . 8.6872635 4 3.443087 7.537786 8 3.828916 9.174337 4 . . 11 4.6119113 8.796979 4 . . 12 . . 6 4.1700788 8.051225 9 4.0522957 7.879379 5 4.2714095 8.519905 12 . 9.397124 4 . . 11 4.350053 9.331104 7 . 7.406225 14 . 8.708834 4 . . 9 4.480106 8.538177 4 . 5.903953 10 5.299745 8.855093 4 . 7.192451 12 . 9.076223 9 5.844003 9.49366 14 . 6.615362 4 . . 11 4.85 8.838325 6 . . 11 . 8.72442 4 . . 9 4.883366 8.6722355 4 . . 10 . 9.284665 6 . 8.402228 11 . 8.71677 6 . . 9 5.358942 9.384294 5 . 7.656807 11 . 8.825084 4 . . 10 . 8.843168 6 . 6.83704 12 4.831026 8.485221 7 4.805626 7.901927 12 . 8.381459 8 . 2.904088 14 4.1092334 8.096209 9 . . 13 . . 4 . . 9 5.691413 9.402679 4 . 7.549126 10 5.69889 9.436436 5 . . 12 . 8.956568 4 . . 11 . 8.6546545 5 . . 10 . 8.685721 4 . . 9 5.848364 9.711116 4 . . 9 5.524533 9.720165 6 . . 12 . 8.857048 4 . . 10 5.50678 9.239161 4 . . 11 5.623512 9.302464 4 3.9960136 7.908037 10 . 9.298968 4 . . 8 5.209748 8.729355 7 4.3242292 8.236253 13 . 9.090649 5 . . 9 5.301348 9.136825 4 . . 11 . 8.591593 5 . . 10 4.512572 8.920359 4 3.903019 7.232786 8 . 8.60528 5 3.952956 7.508304 12 3.963319 7.887486 5 4.1558967 8.06792 9 4.6119113 9.315055 5 . . 10 4.889866 8.459979 4 . . 10 . 8.583259 5 . . 11 6.492935 10.16804 4 . 6.409692 9 . 8.111072 5 3.364879 7.830787 13 . 8.675574 4 . . 13 . 8.6175995 4 . 5.010635 10 4.3231745 8.798184 7 . 6.152733 14 . 8.339564 4 . . 10 4.5135283 8.639608 5 3.4301834 7.203483 10 4.020402 8.484205 6 . 6.59297 12 4.0716953 8.889424 6 . . 12 5.473197 9.283535 5 . . 10 2.567618 6.998435 6 4.6941495 8.304455 11 . 9.062745 4 . . 9 4.926992 8.887602 10 . 8.016944 14 . 6.620882 6 4.1419077 8.019788 13 4.789593 8.829519 7 4.069587 8.310491 14 . . 4 . . 11 5.058198 8.830907 5 . . 9 5.234525 9.146548 5 . . 11 4.78181 9.088917 5 . . 10 4.290385 8.43055 7 4.582521 9.248757 14 4.495179 8.7043705 6 . 7.963516 10 5.206676 8.832919 9 . . 14 . 8.598786 4 . . 14 . 9.0645 4 . . 9 5.835284 9.541452 8 5.377824 9.080057 13 . 8.650215 4 2.538491 6.815157 9 . 9.120295 7 . 8.165836 13 . 8.934477 6 5.985106 10.340588 11 . 8.657932 4 . . 10 4.2352524 8.363701 5 . 8.93542 13 3.4547815 6.549454 7 5.914371 9.425598 12 . 8.942617 10 . 9.0603895 14 . 8.649444 4 . . 9 6.27616 10.223534 4 . . 11 . 9.738289 5 3.753618 8.309423 10 5.419769 9.14382 6 4.325815 8.351167 10 5.268691 9.334165 4 . . 9 5.161271 9.3525 6 3.4793975 8.119007 12 . 8.789471 end
Code:
. drop if (incom_a==. | cincom_a==.)
(6,266 observations deleted)
. display _N
9190
. bidensity cincom_a incom_a, levels(7) saving(bidensity) replace
file bidensity.dta saved
. use bidensity.dta, clear
. sum _d,detail
_d
-------------------------------------------------------------
Percentiles Smallest
1% 0 0
5% 0 0
10% 0 0 Obs 2,500
25% 0 0 Sum of Wgt. 2,500
50% 0 Mean .0045984
Largest Std. Dev. .0158344
75% .0010251 .1549294
90% .0103232 .1617224 Variance .0002507
95% .0266937 .1621767 Skewness 5.911845
99% .0865262 .2124513 Kurtosis 46.87279The confusion might be stemming from my lack of understanding of kernel density estimation. But at this point I am lost in the documentation and even the slightest guidance would be immensely beneficial.
The graph I want to acquire is something like that
Array
Source. Bjorklund, Anders & Jäntti, Markus. (2012). Intergenerational Income Mobility and the Role of Family Background. Oxford Handbook of Economic Inequality. 10.1093/oxfordhb/9780199606061.013.0020.
Many thanks for your help!
0 Response to bidensity assigns too many 0
Post a Comment