PaleoMag 3.1d35 usage notes

The notes below are intended as a quick and dirty intro to the way the new version of PaleoMag is developing and don't really substitute for a user's guide. Note that the Command key is also called the apple key (PC equivalent is usually control). A couple FAQs at the bottom.

Opening a locality or sample: Choose "Open" from the File menu and select the file. Right now things other than a sample, .SAM, .APP, .DAT, or .LSQ file are a bit unpredictable--feel free to try them and complain (or not). .JRA files now will open (but prompt for missing info). You can open 2G binary data files (though it will first open a list of all the samples, and you double click on a sample name to get that one--will get things so the file originally openned gets openned right away). You can also drag files onto PaleoMag's icon, or double-click files made by PaleoMag (those with a PaleoMag document icon). PC version will quiz you to be sure that the files you choose should be used properly. You can edit some locality information by choosing "Edit Locality" from the Edit menu; you can then click on information at the top of the page or the strat levels and change them.

Opening a sample: Easiest from the Locality List window; double click on the sample. Can also open from the Open... dialog (except for APP format, which lacks separate files for each sample, and .DAT from 2G, which will get fixed). This opens the Sample List Window. Missing fields (such as the core coordinate directions, the tilt-corrected directions) are rebuilt if absent from the relevant orientation information (e.g., core plate orientation and bedding strike and dip).

Opening a means file: Choose "Open" from the File menu and select the file. For a new means file, select "New Means File" from the File menu.

Locality List window: You can sort the samples by name or strat level, though at the moment code is not building strat levels from individual sample files. Double click on a sample to open it. Choose "Export..." from the File menu to save all the data in the locality to a text file with either a tab, space, or comma separating the fields. If Edit:Locality has been chosen, you can click on the fields at the top to change them; you can also change strat levels.

Sample List Window: Select multiple steps by click-drag or using shift-click. The little disclosure triangle will show the sample's information when clicked (the old info window no longer exists). You can make steps "hidden" (points will not appear on plots) using the "Hide Unselected" item in the Sample menu; these then are italicized in the list window. These points can be made visible by control-clicking on them or by "Show All" in the Sample menu. Typing letters will also select or deselect steps; dashes permit a range of steps to be selected. Least Squares fits can be made from the Sample menu or typing command-L. You can change the sample for this window using the "Next Sample" and "Previous Sample" commands. Note that the sorting order in the Locality List window determines what is next or previous. From this window you can export the sample data to a flat text file, choosing which fields you wish to include in the output file (Select "Export..." from the File menu). You can cut or copy selected steps. Note that cutting steps will permanently remove them from the file! You can also paste steps into the list window; the geographic and tilt-corrected directions are recalculated for the new sample from the core direction that is being pasted in.

Editing a Sample: This is one of the only commands to allow you change input data, so use this with great caution, preferably after backing up your data. From the Sample List Window, select Edit Sample... from the Edit menu. On the left are the fields that can be changed (the core plate orientation, the bedding orientation) along with the original values. On the right are previews of the revised data: an equal area plot of the coordinates chosen from the pop-up at the top (tilt, geographic, and core coordinates) and a list of the directions after the editing. The previews are live as you change values. You can choose to fix either the core coordinates, the geographic coordinates, or the tilt-corrected coordinates from the popup on the left side. You can view bedding strike and core plate strike as either relative to magnetic north or true north from the checkbox on the left. Your original data is unaffected until you save the list window from the File:Save command or when prompted when closing (and even then there is a copy saved in the data directory of the original data file); cancel leaves everything unchanged.

Making Least Squares fits: Available from the Orthographic plot window,Sample List window, and Sample Equal Area window OR open the "Current Fit" window and you can see your desired fit type updated as you select/deselect points. [Possible bug/limitation is that the arc constraints from a plane fit not fixed to the origin are calculated from the origin instead of the centroid of the plane fit. Another is that the arc constraints on a plane fit will always be on the side with less than 180° of arc; fits to more than 180° of arc will have the wrong part chosen].

Orthographic (Zijder) Window: Open this from the Sample List Window, J/J0 Window, or Sample Equal Area window. You can select point by clicking on them (note that the List window will update immediately). Points are now highlighted when a click would select/deselect them (lower left corner shows the id of the current point or the scale if no point under the cursor); note that *both* points for a given step are highlighted. Control of the view and coordinates is handled in the popup menu in the window. To zoom in, hold down the option (shift on PC) key and click-drag over the region to zoom (cursor becomes a magnifying glass). To unzoom, double click within the view (might have to be fast double clicking on PC). You can save graphics as a number of raster formats ("Save As.." in File menu) and you can drag-and-drop the picture to other apps or the desktop. To make a file you can edit in another program, print the image but select "To File" in the destination pop-up in the upper right (for a Mac OS 9 Laserwriter 8 driver) or "Print to pdf" (in various places in MacOS X print dialogs) and then choose a flavor of EPS or PostScript your graphics program can deal with (Canvas and Adobe Illustrator will work with regular Mac EPS and pdf files). You can also print; changing the zoom changes the size of the points and text (reduction makes type and line thicknesses smaller; yes, a bit counter-intuitive, but I'm not fretting just at the moment). (Printing a zoomed view will probably not print a page of just the zoomed area--some clipping needs to be established). Labels can be shown and symbol sizes changed (go to the Preferences window).

Sample Equal Area (EQarea) Window: Open from the Sample List Window, J/J0 Window, or Orthographic window. Behaves much like the Orthographic window. You can see the modern dipole from the command in the Statistics menu (and set its color, shape, and whether it is rotated in tilt-corrected directions in preferences)

Sample J/J0 Plot Window: Open from the Sample List Window, Sample Equal Area Window, or Orthographic Window. Behaves similar to the Orthographic and Equal Area windows. When points are hidden, the derivative plot (shaded rectangles) ignores the hidden data. Height of scale is given in info pane at lower left. No derivative is shown between different demag types. Component J/J0 is done by solving for vectors projected into component space. Components need not be orthogonal but should not be degenerate; if close to one another, plot will tend to be erratic. Click on Components button to select number of components to solve and whether they are in geographic ("Geo") or stratigraphic ("Tilt") coordinates. Copying a plot will also copy the data (pasting into a text editor like Simple Text will show this). Dragging a pane from this plot will drop the image and the data (so you get the image in a graphics application or a text table in a text editor). Text copied or pasted is a tab-delimited table with one line of headers.

Least Square Fits List Window: Accessible from the Windows menu or by opening a least squares fits (.LSQ) file. Can sort on any field. Cannot scroll horizontally. Changes (e.g., new fits) are not saved until you save them, either from the File:Save menu or when the window is closed. New fits are put at the bottom until you resort by clicking on one of the header titles. You are prompted to save the file when you close it or quit the program. (The previous file is renamed in the same directory/folder). Arc extents (used for certain statistical tests) are now calculated both when making new fits and when old fits lacking them are read (in some cases recalculations fail on old fits, either when the original sample data is missing, the summary of points used is incomplete, or there is a discrepancy between the least-squares fit and the sample data. Warnings are issued if this happens). Multiple directions can be selected (on a Mac, shift-click-drag selects a contiguous block; command-click allows discontinuous directions to be selected). Double-clicking on a line will bring up the sample's list window with the proper points selected; if that sample already had open windows, all are brought forward.

Least Squares Equal Area Window: Accessible from the Windows menu when a Least Squares Fits List window is open. Plane fits are shown as circles, line fits as squares, and lines derived from planes (via Fisher statistics) are open boxes. Shows selected measurements in different colors than unselected measurements. Otherwise behaves as the Orthographic Plot. Fisher and Bingham statistics available--statistic info is shown if you select Fisher Stats or Bingham Stats from the Statistics menu. [Bug--many times the line with the statistics hides; close and open the disclosure triangle just above the statistics to make it reappear, or double-click on the blank panel to the right of the disclosure triangle (or select Summary from the Statistics menu) to bring up the full summary]. Statistics summary can be hidden by clicking on the disclosure triangle. Type of Bingham or Fisher statistic chosen in pulldown menu in the statistics pane of the window. Statistics update as points are selected/deselected in either this or the Least Squares Fit window. Planes can also have the arc of the plane shown from the second pop-up menu. Double clicking on the panel with the statistics or selecting Statistics Detail from the Statistics menu brings up the Statistics Detail Window; statistics info can be copied to the clipboard from here or dragged as a clipping. Mean directions can be saved to any open means file from this window. For the Fisher Line+Plane or Line+Arc statistics, the directions inferred for each plane fit are shown in the Statistics Detail Window in the format of LSQ data; these can be saved to the .LSQ file. Note that currently the usage calculates the optimal direction separately for geographic and stratigraphic coordinates; hence it is possible (indeed likely) that the geographic and straitigraphic directions of these points will not be related by the tilt of the strata. A reversal test on selected points is available from the Statistics menu. Plane and circle fits are converted to points for a reversal test using the same rule as chosen by the user in the equal area plot.

Important Statistics Notes: The Fisher Line+Arc fits and Line+Plane fits seem to generate slightly different fits than the old 2.3.1 code did. I suspect the error was in the old code and not this, but I would appreciate knowing of any tests. The Bingham distribution still yields rotten error ellipses for girdle data owing to limitations in the lookup tables used (this was true in 2.3.1); thus the error ellipses for lines fit with Bingham As Lines will be much better than (and differ from) those made with the Bingham Line + Plane statistic.

Least Squares Stratigraphic Plot: Accessible from the Windows menu when the Least Squares Fits List is open. Line fits are shown as filled squares, poles to plane or circle fits as circles, and lines from plane or circle fits as open squares. Each pane can be set to a different option: Declination, Inclination, MAD (Maximum Angle of Deviation from LSQ fit), VGP Latitude, VGP Longitude, Angle from Mean, and Angle [from Mean] as Bar. The Angle from Mean options calculate the Fisher Line+Arc autoreverse mean from the currently selected directions (which must be selected in the LSQ List Window or the LSQ EQArea window); the mean is reversed so that a northern hemisphere VGP is normal ("Angle from Mean" of 0) . The "as Bar" version colors the background with black for angles less than 90° and white for angles greater than 90°. Also two choices on displaying plane and circle fits: as the poles to those fits (as circles or dreadful heavy boxes when the graphics are fouled up), or as the point along the arc on each fit closest to the mean direction or its antipode (as open squares). Dragging a pane works like J/J0 plot, where dragging to a text window will copy the x-y data of the plot, and dragging to a graphics window will copy the graphic at the scale on the screen.

Locality Equal Area Plot Window: Displays all the points at a specified demag level from the righthand pulldown menu at the window top. Cursor placed over a point will reveal the sample name at the lower left of the window. Click on the Play button in upper left to cycle through all the steps in the locality; click on the Pause button to stop.

Means Window: Presently you can open an old PaleoMag means file with the Open... command. You can save fits from any open LSQ equal area window to any open Means window, and the means file can be saved from PaleoMag. Clicking on a direction selects/deselects it. Clicking on the two color swatches will change the color of the mean and its uncertainty on a selected LSQ Equal Area plot. The popup menu controls the symbol with which the mean direction will be plotted. The popup at the bottom of the window controls which LSQ equal area will plot the selected mean directions. You can now append directions to a Means file through the Edit menu (Add Mean Direction command); these can be either a direction or a pole position.

Preferences: Preferences window will open with the default colors, labels, and shapes for the frontmost window (global prefs if the frontmost window is not a plot). Click on a color to change it. You can save either local preferences (saved in the site folder) or global preferences (saved to the system Preferences folder). To make a site revert to global preferences, remove its preference file. Note that LSQ Equal Area plots use the global preferences, not local ones.

About PaleoMag: Has link to PaleoMag 3.1 home page and shortcut for sending email. Also shows the version info.

Windows executable note: We don't do much debugging on Windows machines unless we hear of complaints, so if something doesn't seem to work right on a Windows machine, let us know. Thanks to Joe Kirschvink, PaleoMag runs fully on Windows.

Linux executable note: Even more than the Windows version, we don't have any way of testing it. So if you are trying it out, let us know what happens!

Frequently asked questions (and requests)

How can I get some decent looking output?
This has been a long-term problem. Right now the best way to make output (for publication) is to capture the printer output and edit it in another program; this is particularly true for the high-resolution output now available for stratigraphic, equal-area and orthographic windows. Copy and save both are making bitmaps at the same resolution as the screen (except for the copy special for the strat plot); attempts to copy or save vector information haven't worked to date (from time to time, we'll try again). So, on a Mac, try and print the file (ideally to the highest resolution device you have), but redirect the output to a PDF or PostScript file (in OS X, there is a "Save as PDF" command at the bottom of the print dialog; in OS 9 there is a popup menu allowing you to choose a file as the destination). Then you can edit the PDF or PostScript in programs such as Adobe Illustrator or ACD's Canvas. [PC and Linux users, please let me know of good solutions on your platforms].
It is hard to hide or show points....
Can be, 'tis true. Control-clicking on a point in some windows will make it hide; control clicking in the list window can make a hidden point reappear. And we added an option to hide points with a MAD above a certain level. So it might not be too bad.
I'd really like to edit my data file....
This is generally a BAD idea, but the reality is that, from time to time, you need to do this. So do it carefully and back up the original data files! We've added some tools, but they have some limits. You can edit a locality's info by chosing to "Edit Locality" from the Edit menu when the locality window is frontmost. You can then click on magnetic declination or strat level or locality description or latitude and longitude and change them. You have to explicitly save the changes for them to remain in the future. For samples, there are two options. If you select "Edit Sample..." from the Edit menu when the list window is foremost, you can try out different core orientations (expressed as the CIT style orientation of a core plate) and tilt corrections (just as the usual tilt correction by rotation about the strike). Once these are accepted, they are not saved to disk until the user chooses to save them. The second method is to cut steps from the list window; these can then be pasted as text somewhere else or pasted into a different sample (this is for those occasions when measurement got a sample out of order). When pasting into a different sample, the core coordinates are the same and the geographic and tilt-corrected directions are recalculated from them. Important limitations include that no changes are made to the LSQ file either from magnetic declination or sample orientation or cut and paste to the sample list (i.e., your direction will not be changed and in the case of cut or paste, the original steps used in the fit are now not properly identified). It is best to not have any LSQ fits prior to editing the data. And the magnetic declination isn't really carried through everything properly at all (in some formats, it really doesn't change anything; in others, we should probably be rotating all our directions by some amount).
My format isn't supported...
Ah well. We've gotten a few in, but there seem to always be more! Send us an example with documentation of what the file holds and we can see about it. A more general text-based import option is sort of half done but not a high priority at the moment.
It doesn't work on the PC....
Please tell us, along with the particulars (what windows were open, what format data you were using, how it broke or didn't break). We don't have PCs laying about, we don't use them, so our testing is limited and, as we focus on making changes to the code overall, we don't try and test on the PC. If we know of something specific, we can do more. Sorry. Some things are cosmetic--occasionally a pane doesn't open automatically and you have to hit the disclosure triangle a couple of times to make it show. Popup menus don't popup--you might have to select the current item and then navigate with the arrow key. We've heard of windows appearing at the edge of the screen. Occasionally, it isn't our fault--the RealBasic environment just wasn't quite working right.
It would be great if the code did this....
We try and keep track of some of these ideas (see the 3.1 development page) but occasionally lose track of some. If your idea requires some new code (like, say, a different statistic), it helps if you send a code snippet that shows how to do it. In general, if you can fix it somewhere else (e.g., a graphics code for better looking output), it isn't going to be high on our list. You can also vote for things on our list--though we are mainly responding to our own needs, we do notice when we get kicked repeatedly on something.

Return to PaleoMag main page, v3 development page, or news page

Please send mail to if you encounter any problems or have suggestions.

C. H. Jones | CIRES | Dept. of Geological Sciences | Univ. of Colorado at Boulder

Last Modified April 15, 2006