 |
|
|
| Data Handling |
- Built-in spreadsheet-like data editor; supports cut and paste between applications.
- Circular data types:
- Angles (unidirectional and bidirectional [axial])
- Time of day
- Compass Direction
- Day of Week
- Month of Year
- Week of Year
- Day of Year
- New non-circular data types:
- Date
- Linear Numerical
- Label
- Subgroups
- Flexible data handling; sets of data can be entered as separate columns (the old Oriana ver. 1 method) or can be entered pairwise with other data (linear, dates, labels, other circular data).
- With the paired method a file can also have one or more grouping variables added to allow for analysis of subgroups of data. Subgrouping can also be performed on existing data (such as the date, summarized in different ways).
- Data can be filtered, restricting analyses and graphs to just cases that meet a certain criterion (e.g. Angle > 90 or Month = June).
- Circular data can be entered as x/y coordinates defining the beginning and ending point of a vector. They can then be converted to vectors (angle and length) for analysis.
- Special frequency editing mode allows data to be entered in the form of a frequency table.
- Variables can be automatically filled with identical, incrementing or random (uniform or von Mises) data.
- An optional file creation wizard aids in the process of setting up new data files.
- Dates can be summarized into day of week, month of year, week of year or day of year.
- A Find and Replace facility lets you search for particular data points and make global changes.
- Data can be sorted.
- Grouped circular data (data recorded approximately to the nearest interval, such as every 20° or every hour, rather than exactly) are automatically detected and flagged (with user confirmation) so that statistical analyses are adjusted accordingly, where relevant.
- Missing data capability with pairwise deletion where required.
- Import and export from/to Excel, Lotus 123, Quattro, dBase, Paradox, ASCII text.
- Number of cases and variables limited only by available memory and disk space.
|
|
| Statistics |
- Basic statistics
- circular mean
- length of mean vector (r)
- median angle
- circular variance
- standard deviation and standard error
- 95% and 99% confidence interval for the mean vector
- mean weighted by linear variable, plus basic statistics for the linear variable itself
- One-sample tests
- Rayleigh's Test of Uniformity
- Rao's Spacing Test
- Chi-Squared
- V-Test
- Watson's U²
- Kuiper's.
The latter two can be used for tests against both the uniform and von Mises distribution.
- Two-sample tests
- Watson-Wheeler F-test
- Chi-squared test
- Mardia-Watson-Wheeler test
- Watson's U² test.
- Multisample tests - The first three two-sample tests above can also be performed as multi-sample tests.
- Circular-circular and circular-linear correlations.
- All analyses can be performed on subgroups of data as well as the whole data set.
|
|
| Graphics |
- Graph types include
- Rose diagrams
- Circular histograms
- Linear histograms
- Arrow graphs (like a histogram, but an arrow is used for each class rather than a bar or wedge)
- Raw data plots
- Distribution graphs (plot data against expected distribution for either uniform or von Mises distributions)
- Stacked and stepped circular histograms/rose diagrams (bars or wedges divided into segments to reflect the proportions of each subgroup within the data)
- Stacked raw data plots (points representing each data point are colored according to their subgroup)
- Two-variable circular histograms (segments of each bar represents the classes of a second circular or linear variable e.g wind speed plotted as segments on bars representing wind direction)
- Circular-linear plots (these incorporate linear data along with your circular data. An arrow, bar or wedge is drawn for each case, with the length representing the linear data)
- Circular scatterplots (these also plot linear and circular data, with each data pair being plotted as points on a circular plot)
- X/Y Scatterplots - can plot any circular or linear variable on
- X/Y Cartesian axes.
- Q-Q plot - compare distributions of two circular samples to assess whether they have the same distribution
- Mean and confidence limits can be shown on histograms. Mean can alternatively be represented by a mean vector, where the length of the arrow reflects the mean vector length (r).
- With paired vector data (consisting of angle and length) the mean vector can be weighted by the vector lengths.
- Axes can be extensively customized with regard to number of grid lines, scaling (linear, percentage or logarithmic), format and orientation of labels.
- Circular graphs can be drawn with the 0°/N point at user chosen position on the circle, and oriented either clockwise or counter-clockwise.
- Division of segments in two-variable histograms can be customized or manually defined.
- Bar width of histograms and rose diagrams can be changed.
- Rose diagrams can have frequency depicted by either radius or area of wedge.
- Linear histograms can be plotted as double width plots to emphasise cyclicity in data.
- Table of frequencies used to produce histograms can be displayed in the Results window.
- Graph types and options can be modified without having to create a new graph window.
- Graph appearance can be customized (fonts, color of axes, thickness of lines and grids, color of background of various sections, series of colors, patterns and symbols for stacked and subgrouped diagrams).
|
|
|
| |
