The Data Scan Demonstration requires a Macintosh with at least 512K of RAM and at least 400K of disk space. If you are using a 512K Enhanced machine, be sure the RAM cache is off. Data Scan requires System 4.2 (or later) and Finder 6.0 (or later).
Demonstration Disk Limitations
This demonstration version of Data Scan processes MacPaint files only and will not allow the extracted coordinate data points to be saved or printed. In order to view the extracted coordinate data, you will need to use the Modify Data function, which is described below.
Please call BrainPower, Inc. at (818) 884-6911 if you have questions about Data Scan.
Starting Data Scan
If you have one disk drive, you will need to transfer all the files on the Data Scan disk to a system disk. If you have two disks, insert your system disk into the internal drive and Data Scan into the external drive. If you have a hard disk, copy all the files from the Data Scan disk into a folder on your hard disk. To start Data Scan, select the Data Scan icon and choose Open from the File menu.
Overview
Opening a MacPaint file in Data Scan prompts the program to read the file and locate the x and y–axes of the graph. After the file is opened, specify information about the graph (the chart type, the plot area coordinates, and desired processing operations such as smoothing or interpolation) and prompt Data Scan to perform a processing pass extracting coordinate data from the graph. Specifications can be modified and additional passes performed.
An Example
The Opening Pass
•> Choose Open Bitmap... from the File menu and the standard Macintosh open–file dialog box appears.
•> Double–click on the folder titled Example and the file in that folder is displayed.
•> Double–click on the file titled Chemical Spectrum.
Data Scan performs the opening pass over the graph. A dialog box appears monitoring the progress of the pass. After the pass is complete, the graph window opens presenting the initial view of the graph. Over this window, the Chart Type dialog box appears.
Preparing the Processing Pass
After the opening pass, Data Scan has no concrete information regarding:
• the type of chart being analyzed,
• the range of data represented on the chart,
• extraneous markings on the chart, or
• the quality of the bitmap rendering of the chart.
This information is specified before the processing pass is performed.
Chart Type
Data Scan has properly diagnosed the graph as a line chart (Line chart is selected in the Chart Type dialog box). No additional information needs to be specified in the Chart Type dialog box for this graph. The graph does not have logarithmic axes, it does not depict multiple curves, and it does not have either Y or X–grid lines.
•> Click the close box in the Chart Type dialog box and the box closes.
Data Range
Data Scan entered default axis ranges during the opening pass based on the number of pixels on each axis. Correct values for the origin and the extremes of the x and y–axes should be specified before the processing pass.
•> Click on the value 397 at the top of the y–axis and a small data entry window appears above the value.
•> Enter the value 9000.0 and click in the graph window. The entry box disappears leaving the value on–screen unchanged. However, after the processing pass, Data Scan sets the axis ranges to the values specified.
•> Click on the value 564 at the end of the x–axis and a small data entry window appears above the value.
•> Enter the value 2000 and click in the graph window.
•> Click on the value 0 at the beginning of the x–axis and a small data entry window appears above the value.
•> Enter the value 350 and click in the graph window.
•> Choose Process Bitmap from the Filter menu.
A dialog box appears monitoring the progress of the processing pass. After the pass is complete, the chart window redraws. Notice the specified range limits now appear on the graph. Data Scan has now determined numeric values for each point on the graph based on the ranges specified.
Data Editing
Data Scan allows you to edit the extracted data. This is useful when imperfections in the scanned file are interpreted as data.
•> Choose Modify Data from the Edit menu. A scroll bar appears at the bottom of the graph window. To the left of the scroll bar, the x and y values for a point are displayed.
•> Click on the right arrow of the scroll bar several times. Vernier lines extending from the x–axis and y–axis appear marking a point on the graph. The point's coordinates are reported on–screen to the left of the scroll bar. Clicking the scroll–bar arrows moves the verniers one data point. The newly marked point's coordinates are displayed.
Imagine that the data causing the second large peak in the chart (the highest peak) has been incorrectly extracted due to a smudge on the original hard–copy graph. (You have determined this by comparing the hard–copy with Data Scan's analysis, but were unable to make the correction on the hard–copy.)
•> Scroll to this peak using the verniers as a guide (dragging the elevator in the scroll bar allows you to traverse many points quickly). The peak begins with the point (815.2, 7144).
•> Choose Clear Point from the Edit menu.
The point's coordinates change from bold type to regular type indicating that the point is marked for deletion. (The point is not actually deleted from the data until a processing pass is performed.) Also, after you scroll off the point, the vernier bars marking the point are retained on screen indicating visually that the point is marked to be removed.
•> Click the right scroll bar arrow once. The point selected is (818.1, 8436).
•> Type Command/B (the keyboard equivalent for choosing Clear Point).
•> Click the right scroll bar arrow once. The point selected is (821.0, 8470).
•> Type Command/B.
•> Click the right scroll bar arrow once. The point selected is (823.9, 7858).
•> Type Command/B.
•> Click the right scroll bar arrow once. The point selected is (826.9, 7054).
•> Type Command/B.
•> Click the right scroll bar arrow once. The point selected is (829.8, 6351).
•> Type Command/B.
•> Choose Process Bitmap from the Filter menu. A pass is completed and the peak is removed.
• Click on the arrows in the scroll bar and scroll the vernier lines over the area where the peak was removed.
The verniers still trace the removed points and the point's coordinates are reported in regular type in the normal place at the bottom of the graph window. You can always see what points have been removed using this tool.
Interpolation
During the opening pass, Data Scan extracted 563 points to represent the chart. We have removed 6 points leaving 557. (This number appears at the top of the chart.) However, you can select any number of points fewer than the original number (563) to represent the chart.
•> Click on the number 557 in the label 557 data points at the top of the chart window, and a small data entry window appears above the value.
•> Enter the value 300 and click in the graph window. (As with axis ranges and deleted points, the number of points reported on–screen does not change and interpolation is not performed until a processing pass is performed.)
•> Choose Process Bitmap from the Filter menu. The dialog box monitoring the pass appears. After the pass is complete, the graph window opens again.
The new graph window depicts the original line chart (comprised of 563 points) and the new line chart comprised of 300 points. The points of the new chart are drawn as boxes. This view allows you to determine if the new points accurately reflect the chart. In this instance, there is virtually no loss of detail suffered by depicting the chart with only 300 points.
Data Scan's Menus
The Data Scan menu choices are described below in the order they appear in the program. Many choices have keyboard Command–key equivalents. The equivalents are presented to the right of the menu choices on screen.
File Menu
The File menu allows you to open graph files.
Open Bitmap - Opens the standard Macintosh file–selection dialog box. The file you select is opened and the opening pass is performed. A message window appears monitoring the progress of the file conversion. Immediately after the opening pass, a number appears in the lower right corner of the message window indicating the tilt of the graphic (in pixels) when it was scanned. If the graphic was aligned perfectly during scanning, the number is zero (0). Data Scan automatically corrects for small (less than 10 pixels) amounts of tilt.
Quit - Returns you to the Finder.
Edit Menu
The Edit menu choices Undo, Cut, Copy, and Paste have no function in Data Scan. They are included for use with desk accessories.
Modify Data - Allows you to remove coordinate data points from extracted data. Choosing Modify Data causes vernier lines to appear in the graph window marking a data point. A scroll bar (with arrows and an elevator) appears at the bottom of the graph window. The coordinates of the point marked by the vernier lines are displayed to the left of the scroll bar. Using the scroll bar, select any point on the graph. When a point is marked and Clear Point is selected from the Edit menu, the point is marked for deletion. The next time a processing pass is initiated, the point, and any other points cleared, are removed. At least one processing pass must have been completed on the data before Modify Data can be selected.
Clear Point - (See description of Modify Data immediately above.)
Filter Menu
The Filter menu controls the processing of the digitized graphic and the manipulation of the extracted data before it is written to disk.
Set Chart Type - Presents a dialog box with four graph types (Line, Area, Bar, and Point), as well as the following graph specifications: X–Grid, Y–Grid, Log X, Log Y, and Multiple Curves. Selections are made here to conform to the characteristics of the graph being analyzed. Selections are retained from previous analyses (or from use of the Save Setting command), therefore the Chart Type dialog box should be checked before data is saved. (The box automatically opens whenever an opening pass is made unless it has been closed during a prior analysis in the same Data Scan session.)
Filter Options - Presents a dialog box with controls affecting how data is extracted from the graph during processing and how the data is massaged before being saved to disk. The processing controls, on the right side of the box, are Ignore Spaces, Noise Filter, and Line Filter. The data massaging controls are Smoothing, Save Maximin, and Zero Start. Selections made in the Filter Options dialog box are retained from previous analyses conducted during the current Data Scan session (or from use of the Save Setting command), therefore the box should be checked before data is saved.
Ignore Spaces - Allows you to set a discrimination threshold for the interpretation of white spaces embedded in otherwise solid black (data) portions of a bitmap. When Ignore Spaces is selected, vertical spaces in the bitmap equal in pixel size or smaller than the Ignore Spaces setting are interpreted as if they were black (valid data). This control does not affect non-embedded white areas. Ignore Spaces affects the analysis of the file in another very important way. When the control is activated, Data Scan interprets the x–coordinate point (or point set) with the lowest y–coordinate value as data. When Ignore Spaces is not activated, the note Largest Segment appears below the Ignore Spaces control. Data Scan then uses the largest, continuous, vertical string of black pixels (with the same x-coordinate) to compute the y–coordinate data values.
Noise Filter - Causes Data Scan to ignore isolated black pixels or pixel sets of vertical length less than or equal to the setting. These pixels may otherwise be interpreted as data. This filter is helpful if the digitized file has many stray pixels or has reproduced smudges from the original hard–copy graph.
Line Filter -Removes all unbroken solid lines from combination line/point charts. This filter allows the extraction of data plotted with regression, trend, or best–fit lines. Set the Line Filter to a value one greater than the width (measured on the x-axis) in pixels of the widest point symbol. Data Scan then ignores all objects with an unbroken horizontal dimension greater than the control setting. Line Filter is available only for graphs containing point symbols. Selecting Point in the Chart Type dialog box enables the selection of Line Filter (otherwise, the control is dimmed).
Options Help- Presents a table guide to the filter configuration options listed here. Activate Help by clicking the Options Help button in the Options dialog box.
Smoothing- Averages (smooths) extracted data points over a moving average period of width corresponding to the setting of the Smoothing control. For example, the minimum smoothing value of 5 implements a weighted average of the data over a period that is 5 data points wide. Data Scan utilizes the Golay–Savitzky smoothing algorithm. The smoother the curve in the original data, the greater the amount of interpolation that can be performed without introducing excessive distortion. It is usually necessary to experiment to select the best setting in each case. The width of the moving average period is adjusted with the scrolling control below the Smoothing button. Both the width of the smoothing period and the percent of the total data represented by the smoothing period selected are displayed. Consecutive, repeated smoothing of the data have no cumulative effect. Each smoothing operation begins from the original set of data points.
Save Maximin - Saves the minimum and maximum y–coordinate values corresponding to the thickness of the digitized lines (data curves) or points. Save Maximin doubles the size of the output coordinate data file. Two choices under Save Maximin are available through radio button selection.
Save Maximin–Stack - Stack splits each of the originally extracted data points into two points in which:
• each original x–axis coordinate remains the x–axis coordinate for both new points.
• each original y–axis coordinate becomes two coordinates lying equally spaced about the original y–axis coordinate (implicit in the Save Maximin function).
The Stack option shows the two new y–coordinates one on top of the other.
Save Maximin–Stagger - Stagger splits each of the original data points into two points in which:
• each original x–axis coordinate becomes two coordinates lying equally spaced about the original x–axis coordinate.
• each original y–axis coordinate becomes two coordinates lying equally spaced about the original y–axis coordinate (the same as Maximin Stack).
• the smaller of the new x–axis coordinates is associated with the larger of the new y–axis coordinates, and the larger of the x–axis coordinates is associated with the smaller of the y–axis coordinates.
The Stagger option shows the two new y–coordinates diagonally staggered from one and other.
Reverse Data- Reverses the x-axis order of the points. If a chart contains data for the years 1980 to 1986 with time plotted on the x–axis, Reverse Data produces a coordinate data file with descending, rather than ascending, time values.
Zero Start -Determines the coordinate value of the pixel immediately above the x-axis and immediately to the right of the y-axis. By increasing the coordinate values for this pixel, all coordinate values for the graph shift slightly towards the northeast corner of the graph. The default setting assigns the geometric center of the first pixel just inside the plot area as (x=0.5, y=0.5). The Zero Start control allows you to set this pixel as (x=0.0, y=0.0) or (x=1.0, y=1.0). For axes of one pixel thickness, these values carry the following connotations:
(0.0,0.0) The graph's axes surround but do not enter the graph area.
(0.5,0.5) The inside edge of the graph's axes define the outside of the graph area.
(1.0,1.0) The outside edge of the graph's axes define the outside of the graph area.
Process Bitmap - Initiates a processing pass on the graph. A processing pass must be made before data can be modified or saved to disk. Process Bitmap must be selected after new selections are made in the Filter Options or the Chart Type dialog box, or after ranges have been edited, points deleted, or interpolation specified. If you do not enter plot area coordinates before initiating a processing pass, an alert box informs you of this, and offers the choice of continuing with the processing pass (using the default axis boundaries) or cancelling it to enter the plot area coordinates. A message window monitors the progress of the processing pass, just as it does the opening pass.
Save Settings - Saves the current control settings (Chart Type and Filter Options settings, data point delimiter, input file type, and the file type selected on the Scanner menu) in a file called "Data Scan.Settings." This file must be in the same folder as the Data Scan program if these settings are to be used the next time Data Scan is launched. Otherwise, standard default settings apply. (The defaults are: Chart Type:Line Chart, linear axes, no grid, no multiple curves; Options Window: no selections; data delimiter: Tab; and input file type: MacPaint.) If the "Data Scan.Settings" file is in the same folder as Data Scan during program launch, the Open Bitmap... command automatically is selected.
Data Scan is available from BrainPower, Inc. for $199.95
