概述
dstat是一款功能全面的Stata社区命令,用于估计各种汇总统计量和分布函数(Summary statistics and distribution functions),支持复杂样本设计下的标准误差计算,并提供协变量平衡、多变量多子群体同步分析以及影响函数存储等高级功能。dstat=distribution+statistics。
安装
net install dstat, replace from(https://raw.bgithub.xyz/benjann/dstat/main/) net install moremata, replace from(https://raw.bgithub.xyz/benjann/moremata/master/) net install coefplot, replace from(https://raw.bgithub.xyz/benjann/coefplot/master/)
环境
Stata 14 及以上
Stata社区命令:moremata
Stata社区命令:coefplot
示例
dstat summarize 现在新增了 at() 选项,用于计算条件统计量。
. sysuse nlsw88, clear (NLSW, 1988 extract) . dstat summarize (mean median) wage, at(grade=10/18) Conditional statistics Number of obs = 2,244 Condition = grade -------------------------------------------------------------- wage | Coefficient Std. err. [95% conf. interval] -------------+------------------------------------------------ mean | 10 | 4.692721 .2402356 4.221614 5.163829 11 | 5.688235 .3512727 4.999382 6.377089 12 | 6.638048 .1593726 6.325515 6.950581 13 | 8.315217 .4910102 7.352335 9.278099 14 | 9.130599 .4990696 8.151912 10.10929 15 | 9.88578 .6455681 8.619806 11.15175 16 | 9.806044 .4052976 9.011246 10.60084 17 | 10.43081 .4875015 9.474806 11.38681 18 | 11.60784 .50493 10.61766 12.59802 -------------+------------------------------------------------ median | 10 | 4.203828 .2676516 3.678957 4.728698 11 | 5.032206 .230795 4.579612 5.4848 12 | 5.507244 .1082474 5.294968 5.719519 13 | 6.843801 .3397835 6.177478 7.510123 14 | 7.745568 .354447 7.050489 8.440646 15 | 8.904991 .3692172 8.180948 9.629034 16 | 8.542671 .3745201 7.808229 9.277113 17 | 10.06441 .4661785 9.150227 10.9786 18 | 10.59178 .3289128 9.946779 11.23679 --------------------------------------------------------------
当指定了 at() 选项时,twoway dstat 现在支持 dstat summarize 功能。
twoway dstat summarize (mean median) wage, at(grade=10/18) over(union)
新增的 order() 选项可用于改变结果的排列方式。例如,默认情况下,dstat summarize 按变量分组显示结果,而使用 order(statistics) 则会按统计量类型分组显示结果。
. sysuse nlsw88, clear (NLSW, 1988 extract) . dstat summarize (mean median) wage tenure Summary statistics Number of obs = 2,231 -------------------------------------------------------------- | Coefficient Std. err. [95% conf. interval] -------------+------------------------------------------------ wage | mean | 7.792448 .1220428 7.553118 8.031777 median | 6.280193 .0977002 6.0886 6.471786 -------------+------------------------------------------------ tenure | mean | 5.97785 .1166616 5.749073 6.206627 median | 3.833333 .1224775 3.593151 4.073515 -------------------------------------------------------------- . dstat summarize (mean median) wage tenure, order(statistics) Summary statistics Number of obs = 2,231 -------------------------------------------------------------- | Coefficient Std. err. [95% conf. interval] -------------+------------------------------------------------ mean | wage | 7.792448 .1220428 7.553118 8.031777 tenure | 5.97785 .1166616 5.749073 6.206627 -------------+------------------------------------------------ median | wage | 6.280193 .0977002 6.0886 6.471786 tenure | 3.833333 .1224775 3.593151 4.073515 --------------------------------------------------------------
qdef() 选项现在支持更多分位数估计方法,总共17种方法,现在可以通过关键词进行选择。
作者
作者:Ben Jann
机构:University of Bern
邮箱:ben.jann@unibe.ch
主页:https://github.com/benjann/dstat/