Select a catalogue using the "Open" item in the "File" menu (or click on one of the recently accessed file names in the "File" menu). The catalogue must be either a FITS binary table, or any of the other formats readable by the POLPACK package. The catalogue must contain columns containing Cartesian X and Y coordinates for each vector. If the names of these column are not "X" and "Y", then the names should be specified in a text file called $HOME/.polpackrc as in the following example (where the names of the two columns are "PIXEL_X" and "PIXEL_Y"):
% cat $HOME/.polpackrc Column X PIXEL_X Column Y PIXEL_YThis file can also contain the names of other columns; in particular the names of columns containing RA and DEC values can be supplied if these are not just "RA" and "DEC" (the defaults). RA and DEC values are qualified by EPOCH and EQUINOX parameters supplied with the catalogue, but these both default to J2000 if not supplied.
If an image is displayed in the main image window, an attempt will be made to align the vectors with the image. If the image is displaying celestial coordinates and the catalogue contains RA and DEC values, then alignment will be in RA and DEC. Otherwise alignment will be in pixel coordinates (based on the X and Y columns in the catalogue). No other type of coordinate system is supported.
If no image is displayed in the main window, a blank image will be created covering the area of the catalogue. The colour of this image can be controlled as usual using the colour maps and intensity scaling controls in the main image control panel.
The columns to use for the length and orientation of each vector are specified on the "Rendering" tab, and default to "P" and "ANG". This tab also contains controls for setting the appearance of vectors. The appearance of selected vectors and unselected vectors can be controlled separately. Note, entering a blank value in the "Vector scale" box will cause a default vector scale to be calculated and used. If a blank value is entered in the "Thickness" box for selected vectors, then selected vectors are rendered with the same thickness as unselected vectors.
Click on the "Binning" tab, select the required binning parameters and press the "Bin now" button. This will replace the displayed vector map with a map formed by binning all the previously displayed vectors (both selected and unselected). If you do not like the results, press the "Un-do" item in the "Edit" menu to re-instate the original map!
Select the vectors to be removed using any of the techniques in the "How to select vectors" section of this help, and then press the "Cut" item in the "Edit" menu. The selected vectors will be removed from the map. To re-instate the original map, press the "Un-do" button in the "Edit" menu.
Click the "Save as" item in the "File" menu.
Various operations within the toolbox require some vectors to be selected in order for the operation to complete. Vectors can be selected in several ways:
The "Selecting" tab contains a menu labelled "Operation:", which is set to "Select" by default. In this mode, the vectors chosen using the techniques described in the previous section are selected, and all other vectors are de-selected (unless control and shift are pressed, in which case the states of other vectors are left unchanged). If the "Operation:" value is set to "De-select", then this is reversed, i.e. the chosen vectors are de-selected, and all other vectors are selected (unless control and shift are pressed, in which case the states of other vectors are left unchanged).
If "Operation" is set to "Select", then clicking over a blank part of the vector map will de-select all vectors (if "Operation" is set to "De-select", then clicking over a blank part of the vector map will select all vectors).
To remove some vectors from the current selection (i.e. to de-select them), set "Operation" to "De-select" and then use one of the techniques described in the previous section to choose the vectors to be de-selected. Be careful to press control and shift while choosing them, otherwise the sates of all other vectors will be reset to "selected".
The "Integrate" tab displays a single row of values formed by binning all the currently selected vectors into a single vector. This is done by combining the Stokes parameters using the selected method, and recalculating the other columns on the basis of the combined Stokes parameters. The table is updated automatically when new vectors are selected. So to perform simple aperture polarimetry, select all vectors within the aperture, and then read off the integrated values from the table in the "Integrate" tab. No facilities currently exist for removing any "background" Stokes parameters (caused by inter-stellar polarization for instance). Sky subtraction should normally be performed prior to forming the catalogue of Stokes vectors.
Note, if the catalogue contains error information in the form of columns of standard deviations for the Stokes parameters, then there is a limit of 100 on the number of vectors which can be integrated into a single measurement. This limit does not apply though if the Stokes parameters are combined using a simple mean (selected using the "Method" menu on the "Integrate" tab).
In order to use the "Bin" or "Integrate" tabs, assumptions must be made about which catalogue columns contain Stokes parameters. By default, it is assumed that the Stokes parameters are contained in columns names "I", "Q", "U" and "V", with standard deviations in columns "DI", "DQ", "DU" and "DV". To select other columns to use for these quantities, either use the "Column Names" tab, or create a file called $HOME/.polpackrc containing column names prior to using the polarimetry toolbox. This file defines columns as in the following example:
% cat $HOME/.polpackrc Column I STOKES_I Column DI STOKES_DIetc
The second word in each line is the standard column name, and the third word is the column name you wish to use. The standard column names are I, Q, U, V, DI, DQ, DU, DV, P, DP , ANG, DANG, PI, DPI, X, Y, RA and DEC. So the above example uses column "STOKES_I" in place of "I", and "STOKES_DI" in place of "DI". If no entry is included for a given column, then the standard column name is used. If a catalogue does not contain a given parameter, then a blank column name can be given in the .polpackrc file. Thus, the line
Column Vimplies that the catalogue does not contain any V values.
The "Statistics" tab displays a table of statistics for the currently selected vectors. The table is updated automatically when new vectors are selected. These statistics are simple statistics of the values in each column, and do not make any assumptions about the content of each column. For instance, if column "P" contains polarization values, the "mean P" displayed in the table is literally the mean of the P values. It is not the P value implied by the mean of the Stokes vectors. Use the "Integrate" tab for such "Stokes-aware" statistics.
A key to the vector scale can be produced, and is controlled using the controls on the "Key" tab. The position of the key can be changed by clicking and dragging over the key.
A catalogue containing spectropolarimetry data is assumed to have an extra column containing some indication of the spectral channel for each vector (this may be velocity, wavelength, frequency, integer channel number, etc). By default, it is assumed that this column is named "Z", but an alternative column name may be specified by including a line in a text file called $HOME/.polpackrc as in the following example, which indicates that the column named "VEL" should be used as the spectral channel column:
% cat $HOME/.polpackrc Column Z VELOnly a single spectral channel may be displayed at any one time. The spectral channel to use is selected using the "SpecPol" tab. The channel may be specified in two ways:
All subsequent operations act only on the displayed vectors (i.e. vectors for other spectral channels are ignored). Thus, for instance, dragging a box over the map will only select vectors in the specified channel, not in any other channel. If you save or bin the displayed map, the resulting map will contain only vectors for the specified spectral channel.