...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Economic Dynamics and Control >User's guide
【24h】

User's guide

机译：用户手册

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In this short note, we describe the specifications for some of the programs that were used to estimate the models used in the paper "Conditional Volatility, Skewness and Kurtosis: Existence, Persistence, and Comovements" by Jondeau and Rockinger (2003). The programs described in this note are public and may be obtained from the www (http://www.fame.ch/research/papers/OccPapers/Rockinger.htm) or by sending an e-mail (MR@fame.ch) to the authors. All these programs use the optimization algorithm SNOPT developed by Gill, Murray, and Saunders. SNOPT is developed in Fortran 77. The copyright for SNOPT is held jointly by the University of California and Stanford University. The licensing arrangements are taken care of by the Office of Technology Licensing (OTL) at Stanford. We obtained authorization to distribute the models as stand-alone applications. We are also distributing the Fortran programs that may be used as interface for SNOPT, however, for copyright reasons, we cannot distribute SNOPT. Notice that we use a Windows platform and Compac Visual Fortran V6.5.0. The reader of this note may also wish to obtain from the www the reports describing the algorithm implemented in SNOPT. The documentation may be found in Gill et al. (1997b). In the document Gill et al. (1997a), the reader may find information how directions can be given to SNOPT on issues such as convergence criterion. Since convergence does not necessarily mean optimum, the reader may find directions on how to interpret the output of SNOPT. The various stand-alone applications that we provide expect a file with univariate data named SP.txt. If the user wishes to run the programs on his own data, he/she would have to create different univariate datasets. Proceeding in this way is for reasons of simplicity. The applications start by reading in the data provided by the user. They also read in a file with initial values. The user may wish to modify a specification file that instructs SNOPT how to behave. For instancve, one may wish to limit the maximal number of iterations or the precision before convergence occurs.
机译：在此简短说明中，我们描述了一些程序的规范，这些程序用于估计Jondeau和Rockinger（2003）撰写的论文“ Conditional Volatility，偏度和峰度：存在，持久性和联动性”。本说明中描述的程序是公开的，可以从www（http://www.fame.ch/research/papers/OccPapers/Rockinger.htm）或通过发送电子邮件（MR@fame.ch）获得。给作者。所有这些程序都使用Gill，Murray和Saunders开发的优化算法SNOPT。 SNOPT是在Fortran 77中开发的。SNOPT的版权由加利福尼亚大学和斯坦福大学共同持有。许可安排由斯坦福大学的技术许可办公室（OTL）负责。我们获得了将模型作为独立应用程序分发的授权。我们还分发了可用作SNOPT接口的Fortran程序，但是，出于版权原因，我们无法分发SNOPT。注意，我们使用Windows平台和Compac Visual Fortran V6.5.0。本注释的读者也可能希望从www中获得描述SNOPT中实现的算法的报告。该文档可以在Gill等人找到。 （1997b）。在文件Gill等。 （1997a），读者可能会发现有关如何在诸如收敛准则之类的问题上给予SNOPT指导的信息。由于收敛不一定意味着最佳，因此读者可以找到有关如何解释SNOPT输出的指导。我们提供的各种独立应用程序期望一个文件包含一个名为SP.txt的单变量数据。如果用户希望根据自己的数据运行程序，则他/她将不得不创建不同的单变量数据集。这样进行是出于简单的原因。应用程序通过读取用户提供的数据开始。他们还读取具有初始值的文件。用户可能希望修改指示SNOPT如何行为的规范文件。出于示例性考虑，可能希望在发生收敛之前限制最大迭代次数或精度。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号