I want to calculate Gini index using following equation:
Array
where y is teny, ry is rank order, and cov(y, ry) is covariance between y and ry. y_bar is mean of y and N is the number of directors on board. i use following data:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input double(DirectorID CompanyID tend) float teny 855555 2 304 .8345916 1947759 6 1431 3.92862 1291925 6 1614 4.4310226 1475128 6 3865 10.610845 327069 6 5781 15.870968 37576 6 808 2.2182567 341830 6 5145 14.124914 2365698 6 3655 10.034317 341837 6 6635 18.215511 1366174 6 2519 6.91558 543071 6 761 2.0892243 508743 6 1916 5.260124 33553 6 6189 16.991077 1136305 6 1584 4.3486614 1052220 6 5460 14.989705 556521 6 2418 6.638298 446218 6 2192 6.017845 508736 6 4612 12.661633 341813 6 6917 18.989704 86917 6 11138 30.5779 341863 6 3665 10.06177 1891128 6 1278 3.508579 87882 6 4091 11.231297 508723 6 2921 8.0192175 2414353 6 426 1.1695265 333731 7 4383 12.032945 645749 7 3583 9.836651 886566 7 1826 5.013041 1765212 9 153 .4200412 1336543 9 1921 5.27385 2479549 11 1006 2.7618394 1140345 11 974 2.6739876 1538664 11 912 2.503775 1641444 11 365 1.002059 1455539 11 456 1.2518874 605809 11 31 .08510638 333598 12 3495 9.595058 1749602 12 334 .9169527 1009631 12 1299 3.566232 1009630 12 1299 3.566232 1279761 12 621 1.704873 1279758 12 48 .13177763 1124417 12 1078 2.959506 641561 12 940 2.580645 1009636 12 515 1.413864 371952 12 590 1.6197666 180640 12 730 2.004118 1090208 12 731 2.0068634 626790 12 2340 6.424159 34066 12 365 1.002059 1009637 12 2400 6.588881 1402319 12 730 2.004118 1009641 12 1461 4.0109816 551393 12 250 .6863418 1009642 12 822 2.256692 552193 12 2065 5.669183 1202566 12 809 2.221002 540291 12 2617 7.184626 1601634 12 2647 7.266987 626804 12 1460 4.008236 341468 12 979 2.6877146 1515539 12 1461 4.0109816 626679 12 1058 2.9045985 1084239 12 1774 4.870281 626665 12 730 2.004118 451488 12 882 2.421414 1759874 12 730 2.004118 493920 12 1319 3.621139 1009632 12 2340 6.424159 1664742 12 1430 3.925875 1043852 12 700 1.921757 36354 12 1421 3.901167 2349795 12 792 2.1743307 1301180 12 1341 3.6815374 1009768 12 1903 5.224434 1283040 12 606 1.6636925 1009633 12 1432 3.931366 626786 12 2340 6.424159 2270671 12 730 2.004118 1076429 12 502 1.3781743 322452 12 699 1.9190117 1009639 12 1430 3.925875 1080985 12 365 1.002059 1251946 12 244 .6698696 1722392 12 1127 3.094029 1273767 12 650 1.7844887 555070 12 431 1.1832533 1009624 12 273 .7494853 374881 12 2873 7.88744 1693471 12 1581 4.3404255 1790795 12 273 .7494853 641576 12 1179 3.236788 310234 12 1092 2.997941 8066 12 1852 5.08442 1513909 12 700 1.921757 1097600 12 346 .9498971 40911 12 753 2.0672615 626698 12 3226 8.856555 1009770 12 599 1.644475 1863745 12 731 2.0068634 end
I already tried the following:
Code:
ginidesc teny, by(CompanyID) m(a1) gk(a2) too many values r(134);
Code:
sort teny . bys DirectorID: egen t=sum(_n*teny) . bys DirectorID: egen s=sum(teny) . g gini=2*t/(_N*s)-1 -1/_N . sum gini,d gini ------------------------------------------------------------- Percentiles Smallest 1% -.9999993 -.9999993 5% -.9999993 -.9999993 10% -.9999993 -.9999993 Obs 1,416,271 25% -.9999989 -.9999993 Sum of Wgt. 1,416,271 50% -.999998 Mean -.9999975 Largest Std. Dev. 2.79e-06 75% -.999997 -.9999204 90% -.9999956 -.9999204 Variance 7.80e-12 95% -.9999944 -.9999204 Skewness 11.03627 99% -.99999 -.9999204 Kurtosis 205.236 drop t - gini
showing results like distribution is not by CompanyID wise. I also use the Simpson and find results below:
Code:
egen HHtag = tag(CompanyID teny) . egen HHn = total(HHtag), by(CompanyID) . egen alltag = tag(teny) . egen allN = total(alltag) . . gen Simpson = 1 - HHn * (HHn - 1) / (allN * (allN - 1)) . sum Simpson,d Simpson ------------------------------------------------------------- Percentiles Smallest 1% .9980935 .9964796 5% .9993749 .9964796 10% .999809 .9964796 Obs 1,416,271 25% .9999788 .9964796 Sum of Wgt. 1,416,271 50% .9999976 Mean .9999008 Largest Std. Dev. .0003425 75% .9999999 1 90% 1 1 Variance 1.17e-07 95% 1 1 Skewness -5.436333 99% 1 1 Kurtosis 37.92618
0 Response to Gini index
Post a Comment