| SigmaPlot 12 New Features and Improvements
Microsoft Office Style Ribbons
|
|
|
||
|
|
SigmaPlot 12 New Features and Improvements in Detail
Microsoft Office Style Ribbons
Ribbons provide context sensitive feature grouping for enhanced functionality and ease of use. These include high resolution button icons. Pressing the Alt key will display Key Tips for those who use keyboard shortcuts instead of the mouse. Also hot keys are displayed in expanded tooltips. [ top ]
The new graph object and property browser is shown below. Object selection in the graph is displayed in the upper Page Objects window and the object's properties displayed in the Object Properties window below. The user selects an object on the graph which is then displayed in the Page Objects window and its properties are displayed in the Object Properties list below. [ top ]
Upon selecting or changing a property the graph updates immediately with no clicking of Ok or Apply required. This feature is called the Dynamic Update. Property group styles may be copied and applied across like groups thus simplifying property specification. New window show/hide/Auto hide buttons are seen in the upper right corner of the windows above. These are used for window real-estate management. [ top ]
New Tabbed Window Views
New tabbed window views enable quick window management such as window selection, closing, and grouping. This allows you to see all window types in one place and then select the window you want. This replaces the messy collection of windows scattered about your workspace and organizes them logically. Just click on the window you want or scroll right or left to find it. Four tabs for two graph pages and their two worksheets are shown below. You see that the graph is associated with Data 3. So just click on the Data 3 tab to see the data for the graph. [ top ]
New Docking Panel Guides
New docking panel guides provide simple window placement. This includes the Graph Gallery, Templates, Layouts, Object and Properties windows. Simply move the window and cursor over the appropriate guide, release the mouse button and the window pops into place. This significantly improves the previous method which was difficult to use. [ top ]
New Tabbed Groups
Tabbed groups allow you to group open windows either vertically or horizontally. This collects like windows for easy comparison and better window management. Right click on a tab and select one of the two types of tab groups shown below. Then drag and drop tabs onto that group. If you have a collection of graphs that are related then you can group them together. [ top ]
New Mini-Toolbars
Object-specific mini toolbars are displayed adjacent to selected graph, worksheet and report objects. This allows rapid editing of object properties. For example, right click an area in the graph and change the fill color using the paint pot icon. In general the toolbars in SigmaPlot 11 are located around the edges of the window. So modifications require back-and-forth mouse cursor movement. The mini-toolbar provides property selection close to the object of interest. Clicking on the graph brings up the mini-tool bar shown on the left below with bring-to-front/send-to-back, group/ungroup, enter text and color fill for the graph background. [ top ]
|
|
|
|
New Quick Access Toolbar
A customizable Quick Access Toolbar (QAT) for frequently used user-specific controls has been implemented. Most controls in SigmaPlot 11 are located in Graph Properties which requires multiple mouse clicks to access. Placing controls in the SigmaPlot 12 QAT gives immediate access. In the image below the Close All control second from the right will close all windows enabling quick workspace cleanup. The rightmost New Notebook control opens a new notebook. [ top ]
Improved Graph Page Navigation
Multiple Zoom and Pan Controls on the Status Bar
Zoom, pan, +/- controls and zoom dialog & slider are now located on the status bar where they will always be available. The status bar shows the hand cursor for panning followed by the custom zoom magnifying glass icon, then the +/- buttons for 10% zoom changes followed by the zoom slider. The hand cursor allows panning in any direction and activating it with the space bar makes it very easy to move the graph page. The plus/minus buttons give a “digital” control over the zoom level. The status bar below shows these controls. [ top ]
Middle Mouse Wheel Scrolling
The middle mouse wheel now provides quick vertical window scrolling. Just select the window and scroll that window with the mouse wheel. [ top ]
Worksheet Mini-Toolbar
A right-click mini-toolbar is available for worksheet cell editing. This can be used to Cut, Copy, Paste, Delete selected, Select All and Transpose Paste. Its advantage is that it works directly on the worksheet region that you are editing. The following is a view of the mini-toolbar. [ top ]
Zoom Enabled Worksheet
All zoom and the hand cursor pan controls on the status bar, except the Custom Zoom, control the worksheet zoom level. This is a very useful feature since quite frequently you scroll the worksheet when you really want to zoom out and pan. The plus/minus buttons zoom by 10%. The slider and panel control can be used also. Press the space bar to toggle the hand cursor for panning.
Quick Transforms are Saved and Shown in Property Browser
Quick transforms that are run on a worksheet are saved with the worksheet and listed in the Property Browser. They can then be rerun from this list. [ top ]
Middle Mouse Wheel Scrolls Worksheet
The middle mouse wheel can now be used to vertically scroll the worksheet too. This is much simpler than using the slider controls. [ top ]
Formatted Text in Worksheet Cells
Text in the worksheet can now be formatted with super and subscript, bold, Greek letters, etc. and can be used in tick labels. Subscripts are shown in the X axis tick labels in the graph below. [ top ]
Custom Tables in Reports
Tables with predefined styles or user-customized may be placed in the SigmaPlot report. Data in the table may be copy and pasted to or from an Excel worksheet. The control for inserting a pre-defined or custom table is shown below followed by an example table. [ top ]
Zoom Enabled Report
All zoom controls on the status bar, except the custom zoom, determine the zoom level in the report. The hand cursor panning control is also not available in the report. [ top ]
Both Vertical and Horizontal Rulers Exist in Report
The report has been improved by adding the vertical ruler so that both vertical and horizontal rulers are available. [ top ]
Control Report Background Color
The report background color may be changed using the Page Background control on the Format ribbon of the Report tab. [ top ]
PDF Button Added
PDF output is frequently used so a button specifically for this output has been added in the Graph Output ribbon of the Home tab. Also PDF export has been enhanced by the use of a new version of the report engine. [ top ]
Drag and Drop Word Content to Report Page
Word 2007 and 2010 content may now be drag and dropped directly to the report page. [ top ]
Curve Fitting Weight Variables may now be Specified Generally
Weight variables in the curve fitter may now be specified quite generally as functions of the parameters. These weights will change at each iteration of the curve fitter which was not the case in previous versions. As special cases this general specification includes the three new data weighting features:
- weighting by predicted values (a commonly requested and statistically desirable feature)
- weighting as a function of residuals (for robust regression)
- weighting as any function of the parameters (used by those that have measured their experiment errors in detail or know them from other's work)
Weighting by the predicted values is known to result in better curve fit statistics. Robust regression will tend to ignore outlying data values and result in a better fit to the non-outlying data. Some scientists perform replicate measurements to determine what the measurement error distribution is and then create a weighting function. They can then incorporate these errors in the curve fit by using this predetermined weighting function. [ top ]
Parameter covariance matrix and confidence intervals added to nonlinear regression reports
Two statistics have been added to the nonlinear regression report: 1) parameter confidence intervals and 2) the parameter covariance matrix. Both can be used to obtain estimates of the error in the parameters of a curve fit. [ top ]
|
Covariance Matrix: a b a 0.8475 b 0.0255 0.0009 Confidence Intervals: Coefficient 95% Conf-L 95% Conf-U a 21.3497 18.4199 24.2794 b 0.7522 0.6551 0.8492 |
Implicit Function Curve Fitting
Implicit() can be very useful in curve fitting functions where the equation you want to fit is implicit. An example of this occurs in drug synergy problems where the use of one drug causes the second drug to have a more potent effect. The equations for a simple implicit function curve fit are shown below. [ top ]
One-Sample Signed Rank test
This is a new statistic which determines whether the median of a data set is significantly different from a user-specified value. It is used when you want to compare some data to a fixed value and you know the data is not normally distributed. The prior existing parametric one sample t-test is used to determine if the mean value in a column of data was significantly different from a single value entered by the user. [ top ]
Deming Regression
Deming regression allows for errors in both X and Y variables – a technique for method comparison where the X data is from one method and the y data the other. The Deming regression method basically extends the normal linear regression, where the X values are considered to be error-free, to the case where both X and Y (both methods) have error. Hypotheses can then be tested, slope different from 1.0 for example, to determine if the methods are the same. For example, it might be used to compare two instruments designed to measure the same substance or to compare two algorithmic methods of detecting tumors in images. The graph compares the two methods to determine if they are different or the same. A report shown below gives statistical results. [ top ]
|
Simple Deming Regression Data Source: Data 1 in Notebook2 X-data: Column 1 Y-data: Column 2 Standard deviation for each X-measurement: 1.0000 Standard deviation for each Y-measurement: 1.0000 Number of observations 25 Number missing 0 Data correlation coefficient 0.9850 Chi-square statistic 12.0218 Reduced chi-square 0.5227 Degrees of freedom 23 Coefficient Std. Error 95% Conf-L 95% Conf-U Intercept 0.0696 0.4303 -0.8205 0.9597 Slope 1.0114 0.0371 0.9347 1.0881 The standard deviation for each observation is only known up to some common scaling factor. The standard errors for the parameters have been corrected by an estimate of this factor, the square root of the reduced chi-square. Hypothesis Testing: Test for slope = 0 F= 1527.7339 P= <0.0001 Test for slope = 1 F= 0.0973 P= 0.7579 |
Normal distribution comparison
For a preliminary quality control analysis the engineer might collect data and quickly look at it assuming it is normally distributed. SigmaPlot generates normal distribution curves for each data set using the mean and standard deviation of the data. By examining the mean and variance (and other statistics) the engineer can quickly determine if a problem exists. The analysis in SigmaPlot produces a graph for visual interpretation and a report for numerical examination. The graph and report for printer tray gap measurements in the four tray corners is shown below. The normal distributions are compared to the limit lines also graphed and to each other. The data statistics are shown in the report. [ top ]
|
|
|
Parallel line analysis
Parallel line analysis determines if linear regression slopes and intercepts of multiple data sets are significantly different. It is commonly used in the biosciences to determine relative potency (EC50), bioassays for specific coagulation factors and inflammatory lymphokines and for radioimmunoassays for prostaglandins. In the example below the slopes are not different (P > 0.05) but the intercepts are (P < 0.0001). The report is written to describe the results in understandable language. [ top ]
|
|
|
Bland-Altman method comparison technique
The Bland-Altman graph and statistics is another approach to method comparison. The Bland-Altman method is a plot that shows the difference between the two methods and computes the 95% limits of agreement. If the difference between the limits of agreement is small then the two methods agree and are considered to be the same. The graph on the left directly compares the results of measurements using both methods. The graph on the right plots the difference Y-X versus the mean (Y+X)/2 and uses the limits of agreement technique developed by Bland and Altman to determine if the two methods are the same. [ top ]
Improvements in post-hoc testing in ANOVA
The P value for Dunnett's test has been added. It is not found in some statistics programs but is now in SigmaPlot. The computation of multiple comparisons in 3 Way ANOVA has been improved to provide increased accuracy. [ top ]
New Analysis and Transform Language Features
Enzyme Kinetics Analysis
Many reactions are controlled by enzymes. Therefore enzymes can be used as drugs in certain situations. A way to measure the enzyme's behavior is by way of its reaction rate. Enzyme Kinetics Analysis provides the method to analyze reaction rates. This feature provides the curve fitting and graphing capabilities needed to analyze and present enzyme kinetics data. Import the data, select the type of study, select candidate kinetic models from over 40 different equations and select the interactive graphs to display the results. SigmaPlot's powerful non-linear curve-fitter then fits all selected equations to your data and ranks them by goodness of fit. The selected interactive graphs help demonstrate the kinetic mechanism. The graphs shown below are created by the enzyme kinetics analysis feature and, since most are based on ratios of the measured variables, are very difficult to generate by hand. These are just four of the seven graph types provided. [ top ]
Worksheet row and column titles from the transform language
There is now a transform statement, cell(0, col) or cell(row, 0), for inserting column titles into the worksheet. This significantly improves the readability of worksheets. This seemingly minor feature makes transforms more useful since the results can now be annotated with column titles. Thus users can return to a notebook at a later date and determine what created the columns in the worksheet rather than, in some cases, have absolutely no idea. Compare no column titles on the left to those inserted from a transform on the right. [ top ]
|
Rather Than This |
You Have This |
|
|
|
New Root() and Implicit() transform functions
Root() and Implicit() are new transform functions for finding roots of equations. They are used in many disciplines where the equations are not simple. As an example for root() you might want to find the x values for which tan(x)=1. You would use the root() function in the transform language to do this. A transform which uses the root() function to compute the two roots for an ellipse is shown below. [ top ]
|
'Find Roots of Ellipse 'Input a=1 b=.2 c=1 x= data(.1,.9,.1) 'Equation k(v)=a*x^2+b*x*v+c*v^2-1 'Duplicate x values n=size(x) xout=lookup(data(1,2*n)/2, data(1,n), x) 'Output with two roots col(1)=xout col(2)=root(k(v),v,-10,10,2) |
|
Implicit() on the other hand allows you to find y values for functions where you can't write y explicitly in terms of x. For example, y = 2*x is explicit since given x = 3 you can compute directly that y = 6. But y = x + x*tan(y) is implicit in y since, with y on both sides of the equation, given an x you can't find y directly. The implicit() function allows you to find y. [ top ]
Enhanced Graphics Features from GDI+
The improved customizable GDI+ graphics software support provides enhanced graphics features:
- Color gradients are available for almost all objects including area fills and lines. You see color gradients in the popular science journals and in poster sessions at trade shows. They are also frequently used in PowerPoint presentations. Though not absolutely required they provide an artistic look. The following graph has a gray background gradient and a color gradient to white for each bar. [ top ]
- Anti-aliasing minimizes jagged lines. This provides a much more professional look on the computer monitor. The graphs below show antialiasing in use on the left but not on the right. [ top ]
|
|
|
- Area objects may be made transparent with the user setting the degree of transparency. Also, areas with color gradients have two controls for both the initial and final color transparencies. As shown below, transparent symbols can be used to show the “mass” of overlapping objects. Also, transparency can be used to show results that lie behind other results. In the graphs below, compare the transparent symbols on the left to the non-transparent on the right to see the histogram “mass” effect of overlapping symbols. [ top ]
|
|
|
Radar Graph
The radar graph allows plotting and comparison of multiple variables beyond the two or three previously allowed in SigmaPlot. It is not unusual to graph 5 to 20 variables in a radar graph – 35 are shown in the graph below. A new data format was created to allow input of the variable names in a worksheet column. There are six radar subtypes for almost any application: scatter, line, line & scatter, vector, area, line & error band. Two area radar plots of 35 variables are shown below. The areas are transparent (new transparency feature) in order to show all variable values for both plots. [ top ]
Dot density graph.
The dot density graph is a graphical arrangement of data points which is designed to display the density of data. Data is graphed vertically or horizontally as histograms with data points displaced to prevent overlap. Plots look like they have been prepared by an artist with data points either touching or nestled together. Points may be displaced from one another or overlapped to create a grape-bunch effect. SigmaPlot's implementation allows the comparison of multiple dot density plots with additional descriptive plots that include mean lines, median lines, box plots and percentile lines. The graph below shows three dot density plots each with their associated box plot. [ top ]
Previously-set user options will now be maintained when an upgrade or new version is performed. This very frequently requested feature has now been added for Version 12 and higher. It will not be necessary now to reenter user-specific options. [ top ]
The graph page measurement units (“inches”, “millimeters”, “points”) is now available for use in macros. The page measurements feature enables macro writers to make their macros work on different international units.
Also the legend state (on, off) may now be controlled. Controlling the legend state has two uses: 1) preventing the display of a legend for aesthetic reasons and 2) preventing the display of an extremely long legend in situations where this happens (scatter plot using different symbol colors, for example). [ top ]