|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.sao.ru/precise/Midas_doc/doc/94NOV/vol1/node64.html
Дата изменения: Fri Feb 23 14:07:47 1996 Дата индексирования: Tue Oct 2 17:22:14 2012 Кодировка: Поисковые слова: п п п п п п п п |
In this section we describe several of the mathematical operations that can
be performed on table data. More specialised topics are described in
Chapter 
, "Fitting of Data", and in
Chapter 11 Vol B, "Multivariate Analysis Methods".
Arithmetic operations between columns in a table are done with the command
COMPUTE/TABLE. The selection flag is reset by this command. Special functions, named according to the FORTRAN
mathematical library, are also available.
Simple statistical descriptors are displayed with the command STATISTICS/ TABLE. These descriptors are stored in output keywords for further usage in a procedure.
A set of histogram-related commands, with qualifier HISTOGRAM, allow the graphic display of the histogram of a column ( PLOT and OVERPLOT commands), the printout of histogram values ( READ and PRINT commands), or the generation of a 1D image with the histogram of a column (command COMPUTE/HISTOGRAM).
Linear or polynomial fits in one or two dimensions can be performed on table columns with the command REGRESSION, qualifiers LINEAR and POLY respectively. The coefficients and error estimations are kept in output keywords that can be stored as table descriptors with the command SAVE/REGRESSION. Fitted values are calculated with COMPUTE/REGRESSION.
A topic of special interest is the generation of table data from an image
and vice versa. A transformation was already described in
section ,
and consists in copying the data from one format to the other, using the commands
COPY/IT and COPY/TI. Tabular data can be
converted into 1D or 2D image data with the command
CONVERT/TABLE. This command works in several modes controlled by a
parameter. In all cases the sampling domain of the result is defined
by a reference image. The modes currently available are : POLY
(polynomial fit to table data), SPLINE (spline approximation),
PLOT (scattergram of the data in the table) and FREQ
(2D histogram of the data in the table).
For more specific resampling and interpolating algorithms, the commands REBIN and INTERPOLATE will provide full conversion between image and table formats (qualifiers TT, TI, IT and II).
Table: Operations on Table Data