経済学書専門出版 エコノミスト社
Top書籍情報計量経済・統計ソフトウェアGAUSSGAUSS (Windows 95/NT版)
ファイナンス大系
リアルオプション
Excelとその応用
統計学大系
経済学大系
e-ビジネス
計量経済学
ゲーム理論
経済数学
環境問題・環境経済学
人口学講座
ナレッジマネジメント
ビジネス書
NLP理論
複雑系経済学
経営学・商学大系
マーケティング
心理学・行動科学大系
金融工学・数理ファイナンス
マクロ経済学
法情報学
企業法学講座
経営工学大系
計量経済・統計ソフトウェア
オペレーションズ・リサーチ
会計学・簿記・税務
楽しい数学
GAUSS

GAUSS (Windows 95/NT版)

Windows 95 および Windows NT用GAUSS (ネーティヴ32-bit版)
ウインドウズ環境でのマルチタスキングをPC上でGAUSSの全機能を実現


GAUSS is now available in 32-bit native versions for OS/2, Windows 95 and Windows NT. This new version of GAUSS runs in multiple moveable, resizeable windows, and you can:

* Have multiple graphics open at one time
* Cut and paste text between GAUSS command and edit windows, and to and from other programs such as word processors and editors
* Run multiple sessions of GAUSS concurrently
* Run long GAUSS sessions in the background, freeing your computer for other tasks


Pull-down menus simplify configuring fonts, window selection, file selection and other options.


New Data Exchange Facility for Win NT/95

GAUSS for Windows now includes a Data Excahange feature for exchanging data with most Windows spread sheets and data bases. GAUSS matricies and data sets can be exported in a variety of file formats. Files generated by other software can be imported as matricies or data sheets. File formats include: Excel, Lotus, Quatro, Symphony, Dbase, Paradox, FoxPro, Clipper and ASCII.


Legacy DOS applications

GAUSS for Windows can now run your DOS programs in a standard 80 x 25 window. Supports: Foreground/background color at the individual character level ・ ANSI.SYS escape code sequences for color, cursor, and screen control.


On-line Help browser

The browser window extends GAUSS help to include capabilities not previously available, including features to help developers better understand GAUSS code. The new printing capabilities can also be used with the Browser Window.


Graphics

A new graphic utility is provided that extends file export options to include: WMF, BMP, TIFF and CMYK PostScript. Also includes clipboard support for pasting graphics into other applications, and various new color and formatting options.Separate Run and Edit file selection lists

In the top portion of GAUSS's Command and Edit windows are two separate file selection boxes containing lists of the files you are currently running and the files you are editing. With one click you can bring a file in to the editor and with one click you can save it and rerun your main program. This simplification of the edit run cycle when combined with the new make facility makes development efficient, and easier than ever before.

File names can be copied between these lists with one mouse click.


"Make" facility

GAUSS for OS/2 and Windows keeps track of which files have been changed. This eliminates the need for you to keep track of all your changes when editing and running complex programs using multiple files. This also can significantly speed up consecutive runs of complex programs because you no longer need to clear the workspace with the new command each time you make a change to a secondary file.


Foreign Language Interface

Using the Foreign Language Interface is much simpler in this new version. GAUSS for OS/2 and Windows supports any compiler that will generate .dll files.

There are those times when you need a function that GAUSS doesn't have, but for which you already have Fortran or C code available. Incorporating FORTRAN or C in to GAUSS programs used to require "startup" code which is only available with certain compilers. Now all you need is a .dll (dynamically linked library) file, which most UNIX, Windows NT/95 and OS/2 compilers can generate.


Sparse Matrix Routines

Sparse solve and matrix operations in the OS/2 and Windows versions efficiently handle large sparse matrices. GAUSS's CPU storage and times for computing sparse solve and sparse matrix products are on the order of the number of nonzero elements, and therefore are far more efficient than their dense analogues for large matrices.

Functions are included for the following:

* Solve Ax=B with sparse A
* Convert dense to sparse matrix
* Convert packed to sparse matrix
* Check whether x
* Create sparse identity matrix
* Vertically concatenate sparse matrices
* Horizontally concatenate sparse matrices
* Define rows of sparse matrix
* Define columns of sparse matrix
* Return number of nonzero elements in sparse matrix
* Define dense submatrix of sparse matrix
* Define sparse submatrix of sparse matrix
* Matrix product of sparse times dense matrix
* Matrix product of sparse matrix transpose times dense matrix.


Cumulative Distribution Functions

Previously GAUSS users were limited to 3 dimensions with cdftvn( ). Now they are only limited by computational time and RAM. Functions are included for the following:

cdfmvn( ) multivariate Normal cdf

cdfn2( ) compute interval cdf: cdfn(x+dx) - cdfn(x). cdfn2( ) increases accuracy for a common use of cdfn( ). Further gains in accuracy are achieved when taking the log of an interval of the Normal cdf with the lncdfn2( ) function below.

lncdfn( ) log of Normal cdf

lncdfnc( ) log of complement of Normal cdf

lncdfbvn( ) log of bivariate Normal cdf

lncdfmvn( ) log of multivariate Normal cdf

lncdfn2( ) log of Normal interval cdf

The lncdf functions use a special technique to increase the accuracy of the logarithm of the Normal cdfs. This is important for maximum log-likelihood functions which require the log of the Normal cdf. The increase in accuracy can be as much as 10 places in the tails of the Normal cdf.


Other New Features

New Random Number Generators

* Gamma pseudo-random numbers
* Poisson pseudo-random numbers
* Negative binomial pseudo-random numbers
* Beta pseudo-random numbers

New Spline functions

* 2-D spline interpolation under tension
* 1-D smoothing spline under tension

New Loess Regression function computes robust locally weighted regression using the method of William S. Cleveland ("Robust Locally Weighted Regression and Smoothing Scatterplots", JASA, 74:829-836.



Early Access

All of the features and functions in GAUSS-386i (the DOS version) will be incorporated in GAUSS for OS/2 and Windows, including the Source-level Debugger, On-line Help, support for complex numbers, data translation loop, and Publication Quality Graphics.

We are currently shipping an "Early Access" version which does not have the Source-level Debugger implemented. This feature will be in the final release. All customers who purchase the Early Access version will receive a free update to the final release of GAUSS for O/S2 or GAUSS for Windows when it is available.



System Requirements

GAUSS for OS/2 and GAUSS for Windows NT/95 both require a minimum 386 IBM compatible computer with and Intel 387 or compatible math coprocessor, a 486 or Pentium are preferable, a minimum of 12MB of RAM and a hard disk with at least 6 MB of available space.



For a full description of GAUSS you can read " What is GAUSS?".

Also see our Platforms page.