首  页 关于博脉 软件产品 硬件设备 培训课程 主要客户 联系我们
博脉 -- 专业 务实 周到
 
  产品介绍 Introduction
 [lnkBackToProduct]

  产品咨询 Enquiry
若您有任何产品的问题或需求,欢迎随时与我们联络。
我们将很乐意为您服务!
电邮:info@pomine.com
电话:(020) 2264 4217
传真:(020) 2264 4215

您还可以填写产品咨询表:

 产品信息 Products Info 
Stata 9 软件更新 (下载 2006-11-21 补丁)
2006-11-27  pomine.com
Stata 9 可执行文件和 ado 文件都已更新到 2006-11-21 的最新补丁,如您能在 Stata 中直接连接 StataCorp 的服务器,请使用 update all 命令进行自动更新。
www_._p_o__m_i_n__e._co_m
如果不能在 Stata 中直接连接服务器,请参照下面的页面,进行手动更新:
w_w__w__.p__o_m_i__n__e_._c_o__m
Stata 9 可执行文件手动更新说明
Stata 9 ado 文件手动更新说明
www__._pom_in_e_._com_
___2006-11-21 补丁更新内容___
w_ww.po_min_e.co_m_
Ado-files
w_ww.p_o_m__i_n_e.__co__m_
1. bstat exited with an error indicating a name conflict when the stat() option contained a matrix with equation names. This has been fixed.
ww_w._po_min_e_.co_m
2. cnreg, logit, mlogit, ologit, oprobit, probit, regress, and tobit are now noticeably faster (up to five times faster) when used within large loops, such as statsby with datasets that have many small by-groups.
w_ww__._p_o__m_i_n_e_._c_o_m_
3. histogram with an if condition that excluded one of the categories in the by() option resulted in an extra empty plot when option addlabels or kdensity was specified or resulted in an error when option normal was specified. This has been fixed.
w_w_w__.p_o__mi__ne_._c_o__m_
4. nestreg reported confidence intervals for the default confidence level even when a different confidence level was specified in the level() option of the prefixed command. This has been fixed.
w_w__w_.__p_o__m__in_e_._c_o_m__
5. quadchk exited with an error when used to check models using the offset() option. This has been fixed. quadchk also ignored weights and the collinear and constraints() options. This has been fixed.
ww_w_.po_mi_n__e_._c_om_
6. svy now identifies the rare case when a sampling unit bridges adjacent strata in multistage designs. svy treats all sampling units as nested within strata and units from prior stages.
ww_w_._p_o_m_i_n__e._c_o_m_
7. svy: heckman excluded from the estimation sample observations with a missing value in the dependent variable when it should not have. This has been fixed.
w_ww__.__p_o_mi_n_e_.c__o_m_
8. svy jackknife reported incorrect values for the sample size, population size, and number of strata when some of the strata were omitted. This has been fixed.
ww__w__.__p_o_m_i__n_e_.c_o_m_
9. svy linearization is now two to five times faster. svy: mean, svy:
proportion, svy: ratio, and svy: total are considerably faster when the over() option identifies many subpopulations.
w__w_w__.__po_min_e__._c_o__m_
10. svy: mean, svy: proportion, svy: ratio, and svy: total reported missing standard errors due to strata with one sampling unit even when a subpop() option excluded the offending strata. This has been fixed.
w_ww__.__p_om__i__n__e_.__co_m_
11. svy: mean, svy: proportion, svy: ratio, and svy: total failed to omit strata that contain no subpopulation members. This has been fixed.
w_w_w__._po__m_in_e.co_m_
12. svy: tabulate twoway, when option row or column is specified, uses the ratio estimator to compute cell proportions relative to the row or column margins. For direct standardization, option stdize(), a better estimator is available: the standardized ratio estimator. This standardized ratio estimator is now used in place of the simple ratio estimator when stdize() is specified.
ww__w_._po_m__i__n_e_.co_m_
13. svy: tabulate incorrectly reported zero standard errors when used with survey data that had strata with one sampling unit. Standard errors and confidence intervals are now suppressed with a note explaining why when the se or ci option is specified in such cases.
ww__w__._po_m_i_n__e._co_m__
14. xtcloglog and xtpoisson with the re and normal options would ignore constraints on lnsig2u. This has been fixed.
w_ww_._p_om_in_e__.c__o_m_
15. xtintreg, xtlogit, xtprobit, xtpoisson with the re and normal options, and xttobit all accepted the collinear option but exited with an error when used with a constraint to define a collinear model. This is now fixed.
ww_w._p__o__m_i_ne_.c_o_m__
16. xtintreg and xttobit ignored constraints on sigma_u and sigma_e. This has been fixed.
w_ww._p_o__m__i_ne_.__c_o_m_
17. xtline has new i() and t() options. The i() option allows the panels (groups) to be identified using a string variable. The t() option allows a variable with noninteger values and repeated values within group.
ww_w__._pomi_ne.__co_m_
The important effect of this is that xtline can now be used to easily create two-way graphs with separate plots for each level of a categorical variable. For example,
. xtline mpg, t(displacement) i(rep78) overlay creates five line plots of mpg versus displacement, one for each of the five levels of rep78.
www_._p_omine__._c__om__
These plots can be rendered as other twoway plottypes by adding a recast() option. For example, adding recast(scatter) to the command above will produce five scatterplots rather than five line plots.
w_ww__.p_o__m__in__e_._c__o_m_
Stata executable, all platforms
w_ww._p__o__m_i_ne__._c_om_
18. In Mata's ghk() function, when the covariance matrix is not positive definite, the ghk() Cholesky routine will pivot out the offending row and column of the covariance matrix. The ghk() function failed to permute the corresponding upper bound of integration. This has been fixed.
w__w_w.p_omi_n__e__.co_m_
Stata executable, Windows
www_.__po__min_e_._c_o_m__
19. Stata for Windows now supports Microsoft Windows Vista.
w_w__w._p_o_m_in__e__._c_om
Stata executable, Macintosh
w__ww_._p_o_m_ine_.co_m_
20. For efficiency, when resizing a graph window, Stata draws a preview of the graph rather than completely redrawing the graph. However, the preview could be drawn at the wrong size if the window is quickly resized by jerking the mouse. This has been fixed.
w_w_w__._p_om_i_n__e_.__c__om__
21. When there is a -more- condition in the Results window, Stata will continuously poll for a key press to clear the -more- condition, which causes the processor usage to be very high. Stata now does not poll as often when there is a -more- condition, causing the processor usage to be very low.
w_ww_._p__o_m_ine_._com
22. Some graph markers and objects would not be drawn with an outline in the Graph window when using the Quartz 2D graphics engine. This has been fixed. This bug did not affect EPS and PS files.
ww_w_._p__o_m__i__ne__._c__o_m_
23. The console version of Stata/MP for Intel-based Macs could not correctly determine whether an executable update was available. This has been fixed.
ww_w__.p_o_m__i_n_e_.co_m_
24. The Do-file Editor will now treat a tab character as multiple spaces when printing a do-file. The Do-file Editor previously treated a tab character as one space when printing.
ww_w.p_om__i__n__e_.__c_om_
25. Stata now creates the Stata license file as stata.lic instead of Stata.Lic after license initialization. This distinction allows Mac OS X systems using the UFS file system to read the license file because the UFS file system is case sensitive. The default file system for Mac OS X is HFS+ and is not case sensitive, and it will accept both stata.lic and Stata.Lic. Current Stata users who are using the UFS file system can simply rename the file Stata.Lic to stata.lic. Current Stata users who are using the HFS+ file system do not need to rename Stata.Lic to stata.lic.
w_ww_._pomi_n_e_._c_om_
Stata executable, Unix (GUI)
ww__w._pom_i__ne.com__
26. Clicking a link in the Results window to open a help file in the Viewer could cause Stata to crash. This has been fixed.
Copyright ©2003- Guangzhou Pomine Info. Tech. Co., Ltd. 广州博脉信息技术有限公司 ©版权所有 (粤ICP备11101756号-1)