Canoco 5.1 - 生态排序分析软件

Canoco是一套在生态学及几个相关领域内使用排序方法来进行多变量统计分析的常用程序包。 Canoco 5是一套重新重构的Canoco软件，发布于2012年十月，是生态学应用软件中用于约束与非约束排序的工具。

Canoco 5合了排序以及回归和排列方法学，以便得到的生态数据统计模型。Canoco 5线性和曲线单峰方法。使用Canoco 5进行排序，能够了解：

• 生物群落的结构

• 植物与动物群落以及它们的环境之间的联系

•  对环境和(或)其生物群落的假设冲击所能造成的影响

• 在生物群落上进行的複杂生态学和生态毒理学实验的相关处理所能造成的影响

软件

• Canoco 5实现了接近在排版方面的进展，例如变化分区，协同对应分析和基于距离的冗余分析，但主要的进展是用户的友好性。在Canoco 5中，数据导入，分析和绘制图形被到Canoco 5项目中。

• Canoco帮助选择数据转换和分析方法。过去需要多次运行的数值分析，现在可以通过单峰分析模板和分析笔记简洁地总结结果，并允许访问的结果。

• 对一组数据表示所做的分析现在都在Canoco 5项目中手机，共享分析和绘图设置。Canoco 5有助于制作更好的发布质量排序图。

• 手册已经被重新编写，大量的现实生活中的粒子被更新和扩展，以展示处理多变量数据的新方法。

软件功能

• 对两组或三组预测者来说，变化划分很实现，根据部分或非部分分析和使用原始或调整的变化估计计算被解释的变化的各个部分。

• 在变分划分框架中，可以使用邻矩阵主坐标（PCNM）方法。目前的实现符合Legendre&Legendre（2012）中的方法名称（dbMEM）下的建议。

• 计算、测试和绘制主响应曲线（PRC）现在是一项的任务。

• 工队应分析（CoCA，对称形式）可用，蒙特卡罗置换测试两种比较群落类型的共变异。

• 预测器的逐步选择在视觉上得到了增强，现在提供了对I型错误膨胀的保护（用预测器进行初步测试，通过三种方法之一调整P值：错误发现率（FDR）估计、Holm校正和Bonferroin校正）。

• 可以直接测试约束轴，并将两个培训结果与Procrustes分析进行比较

• 您可以地处理物种的功能或导入物种的系统发育相关数据，以及计算和使用功能多样性。

• 可视化功能得到了增强，例如半透明填充色，在纵坐标图中显示变量箭头的校准轴或绘制封闭椭圆以替代封闭多边形。在现有文件格式（PNG、BMP 、EMF、Adobe Illustrator）中添加了JPEG、TIFF和PDF文件格斯的导出类型。

• 您的工作由上下文敏感的帮助系统和Cancona - 系统，帮助您为那您的研究问题Corre选择合适的分析方法。

操作系统：

Canoco在Windows 8 、 8.1或10的32位和64位版本的标准台式机和笔记本电脑上都可以正常工作。

Canoco 5也可以在Microsoft操作系统上运行，从安装了（Service Pcak）2或SP3的Microsoft Windows XP开始。这也Windows Vista和Windows 7。

Canoco 5经过测试，可以在Linux上的Wine环境下运行，并且也可以在CrossOver软件包的环境中运行。

【英文介绍】

Canoco is one of the most popular programs for multivariate statistical analysis using ordination methods in the field of ecology and several related fields. User's Guides of the recent Canoco versions (4.0, 4.5 and 5.0) were cited more than 9200 times in the past 18 years (1999-2017, ISI Web of Knowledge).

Canoco 5 is the latest, much re-worked version of the Canoco software, released in October 2012.

The main features of the Canoco 5 program are summarized in the following points.

Analytical and graphing capabilities are integrated with an easy-to-use spreadsheet data editor in a single program. All analyses done on a set of data tables are now collected within a Canoco 5 project, sharing the analytical and graphing settings.

All statistical methods offered by Canoco for Windows 4.5 are available, such as DCA, CA, CCA, DCCA, PCA, and RDA methods - including their partial variants, with Monte Carlo permutation tests for constrained ordination methods, offering appropriate permutation setup for data coming from non-trivial sampling designs.

For newly available methods see below.

All visualization tools offered by CanoDraw 4.x are available (including loess, GLM and GAM models for the visualization of data attributes in ordination space) and many of them are improved.    

Data can be entered within the program itself or easily imported from Excel (.XLS or .XLSX formats) or from Canoco 4.x data files. Labels no longer need to be shortened to 8 characters, but these brief forms are still available for display in the ordination diagrams and can even be automatically generated from the long ones. Standard coding of factors (categorical predictors) is now used, dummy (0/1) variables are generated internally. The editor allows transformation from dummy variables to factors and, if needed, the reverse.

 Principal coordinate analysis (PCoA) and distance based RDA (db-RDA) are now easily accessible, with new distance measures added (11 distance types in total, including Bray-Curtis, Gower distance, or Jaccard coefficients). Similarly, non-metric multidimensional scaling (nMDS) is also supported.

Variation partitioning is easily accessible for two or three groups of predictors including calculations of individual fractions of explained variation, based either on partial or non-partial analyses and using either raw or adjusted variation estimates.     

Principal coordinates of neighbour matrices (PCNM) method is available within the variation partitioning framework. Present implementation matches the suggestions described in Legendre & Legendre (2012) under an alternative method name (dbMEM).

Computing, testing and graphing of the Principal Response Curves (PRC) is now an easy task.

Co-correspondence analysis (CoCA, symmetric form) is available, including Monte Carlo permutation testing of the covariation among the two compared community types.      

Stepwise selection of predictors was visually enhanced and provides now the support for protection against Type I error inflation (preliminary test with all predictors and the adjustment of p values by one of three methods: false discovery rate (FDR) estimates, Holm correction, and Bonferroni correction.

Straightforward testing of all constrained axes as well as comparing results of two ordinations with Procrustes analysis is available.

You can easily work with species functional traits or import the data on phylogenetic relatedness of species, as well as calculate and use functional diversity.

Visualization capabilities were enhanced with features such as the semi-transparent fill colours, displaying calibration axes for variable' arrows in ordination diagrams, or plotting enclosing ellipses as an alternative to enclosing polygons. Additional types of export in JPEG, TIFF, and PDF file formats were added to existing ones (PNG, BMP, EMF, Adobe Illustrator).

Every step of your work is supported by an context-sensitive help system and by the Canoco Adviser – an expert system that helps you to select a proper analytical method for your research question, correct type of ordination model (linear vs. unimodal), data transformation, or appropriate visualization of the results. It even advises you how to interpret ordination diagrams you create with the help of Graph Wizard.

Advanced users can combine multiple methods in a single analysis, including generalized linear models to correlate scores etc.