This page describes the changes in Release 20.1.
- 4 new directives
- 21 new procedures
- the ability to input, store and display Unicode characters to support languages such as Chinese, Greek and Thai
- ability to add an error bar to any graph (
- spider-web and star plots (
- text progressions (
- input from GRIB2 meteorological data files (
- conversion between latitude and longitude and UK grid references (
- multinomial random numbers (
- double-Poisson probability distribution (
- Cate-Nelson graphical analysis of bivariate data (
- automatic generation of reports from analysis of variance (
- information about aliased terms in analysis of variance (
- screening tests for REML fixed effects (
- indexes to compare groupings from different cluster analyses (
- bootstrap methods to assess reliability of cluster analysis results (
2. What’s new
COMMANDINFORMATION provides information about whether (and how) a command has been implemented.
COUNTER increments a multi-digit counter using non base-10 arithmetic.
TXINTEGERCODES converts textual characters to and from their corresponding integer codes.
TX2VARIATE converts text structures to variates.
ARESULTSUMMARY provides a summary of results from an
A2RESULTSUMMARY provides a summary of results from an analysis by
DCLUSTERLABELS labels clusters in a single-page dendrogram plotted by
DERRORBAR adds error bars to a graph.
DSPIDERWEB displays spider-web and star plots.
FACEXCLUDEUNUSED redefines the levels and labels of a factor to exclude those that are unused.
FALIASTERMS forms information about aliased model terms in analysis of variance.
GBGRIDCONVERSION converts GB grid references to or from latitudes and longitudes or to or from UTM coordinates.
GRIBIMPORT reads data from a GRIB2 meteorological data file, and loads it or converts it to a spreadsheet file.
GRMNOMIAL generates multinomial pseudo-random numbers.
HBOOTSTRAP performs bootstrap analyses to assess the reliability of clusters from hierarchical cluster analysis.
HCOMPAREGROUPINGS compares groupings generated, for example, from cluster analyses.
HFAMALGAMATIONS forms an amalgamations matrix from a minimum spanning tree.
HFCLUSTERS forms a set of clusters from an amalgamations matrix.
HPCLUSTERS prints a set of clusters.
PFACLEVELS prints levels and labels of factors.
PRDOUBLEPOISSON calculates the probability density for the double Poisson distribution.
RCATENELSON performs a Cate-Nelson graphical analysis of bivariate data.
TXPROGRESSION forms a text containing a progression of strings.
VFIXEDTESTS saves fixed tests from a
VSCREEN performs screening tests for fixed terms in a
3. What’s changed
Most of the changes are compatible with Release 19, the previous release. There are a few commands, however, where new options or parameters have been inserted into the existing lists. These may cause problems in statements where option or parameter names have been omitted or abbreviated (see Section 1.7.1 of Part 1 of the Guide to the Genstat Command Language for details). To avoid any difficulty, the name of the option/parameter after the new option/parameter should be given explicitly, and not abbreviated to fewer than four characters.
Any command, where changes in Release 20 may cause incompatibilities in existing programs, is marked in Sections 3.1 and 3.2 by the symbol †. The full details are given in Section 3.4.
FCLASSIFICATION now allows you to suppress the reordering of the terms in the model formula, that usually occurs after a dot, slash or star operator. It can also save the “main term” corresponding to each term in the model. If the term is a pseudo-term, this will be the term to which that term is linked. Otherwise, it will be the term itself.
GET can access the current language settings (specified by
SET), so that authors of procedures can provide output in other languages.
GETATTRIBUTE has a new setting,
coding, of its
ATTRIBUTE option, that returns a logical value to indicate whether a text contains non-ASCII characters, such as Chinese, Korean and Thai characters.
LIST can now save a pointer containing the structures that have been listed.
SET now allows you to specify your preferred language and (optionally) your second choice in case your preferred language is unavailable. This is to allow you to exploit the new ability of some commands to provide output in languages other than English.
VMCOMPARISON can now display the probabilities in a shade plot, with the cells coloured to indicate the significance of the difference between each pair of treatments.
AUDISPLAY now allows you to specify the treatment terms for which predicted means are to be printed.
AYPARALLEL can export the output to a Genstat workbook or an Excel spreadsheet.
CDNROWCOLUMNDESIGN can now provide several randomizations of the best design, for use in different trials.
DECIMALS now uses the “system default” number of significant figures (which can be modified by the
SIGNIFICANTFIGURES option of
SET) as the default for its own
EXPORT can now output non-ASCII characters to text files, in either UTF-8 format or in Unicode. The default for its
MISSING option with csv files is now to leave the cell empty instead of inserting the string
IMPORT can now convert Unicode characters in Excel
xlsx files into UTF-8. Unicode in the column names can either be removed, or used as extra texts, or used as pointer suffixes (default).
IMPORT also now allows for more than one text to indicate missing values, and allows you to supply some of the column names in a text and others in the data file.
RYPARALLEL can export the output to a Genstat workbook or an Excel spreadsheet.
TRELLIS can now plot onto landscape and portrait frames.
UTMCONVERSION can now use the OSGB36 datum (which provides the basis for the UK Ordinance Survey grid references).
VASERIES now allow factors to be specified for their
CHARACTERS can now provide the number of bytes required to store the text. Ordinary characters, like letters and digits, require one byte each. Unicode characters, that are stored in UTF-8 format, may require up to four bytes.
||new default for the