From 83632f0d22dfcc82dfb8d139bf7a37df1f8471a7 Mon Sep 17 00:00:00 2001 From: Thomas White Date: Tue, 28 Feb 2012 16:26:29 +0100 Subject: Tidy up powder_plot man page --- doc/man/powder_plot.1 | 113 ++++++++++++++++++++++++++------------------------ 1 file changed, 59 insertions(+), 54 deletions(-) diff --git a/doc/man/powder_plot.1 b/doc/man/powder_plot.1 index 1588ab04..20018a47 100644 --- a/doc/man/powder_plot.1 +++ b/doc/man/powder_plot.1 @@ -8,82 +8,87 @@ .TH POWDER\_PLOT 1 .SH NAME -powder\_plot \- 1D powder pattern generating program +powder\_plot \- generate 1D powder patterns .SH SYNOPSIS .PP .B powder\_plot --i mypatterns.stream -o mydata.dat [options] +-i mypatterns.{stream,hkl,h5} -o mydata.dat [options] .SH DESCRIPTION -This program takes as input the data stream from "indexamajig" or an hkl file, -or an HDF5 file. It sums the many individual intensities, Bragg peaks or pixel -locations in the HDF5 file, together to form a 1D powder histogram. +This program takes as input the data stream from "indexamajig", an hkl file +or an HDF5 file. It sums the many individual intensities, Bragg peaks or pixels +in the HDF5 file together to form a 1D powder histogram. The output of powder_plot consists of a three line header followed by a tab delimited list of 6 values: - 1/d of the histogram bin where d is the Bragg law d spacing in meters + 1/d of the histogram bin, where d is the Bragg law d spacing in meters the total number of peaks (N) the total intensity in the N peaks the mean intensity of the N peaks - the 1 sigma (standard deviation) of the distribution - the 1 sigma (standard deviation) of the mean of the data + the standard deviation of the distribution + the standard deviation of the mean of the data -Please note that the sigma of the mean is not the sigma of the distribution, -however they are related by the sqrt(N). - -Typical usage is of the form: - -$ powder_plot -i mypatterns.stream -o mydata.dat [options] - -The options include added geometry, beam, PDB files or symmetry elements. The -program will prompt you as needed depending on the file type and data types -selected. +The sigma of the mean is not the same as the sigma of the intensities +themselves. The former quantity measures how accurately the mean intensity has +been determined, whereas the latter quantity measures the spread of the +intensities. .SH HISTOGRAM OPTIONS -You can choose to set the min and maximum 1/d value [units are inverse meters] -with: --min= and --max= The default value is to use the -entire detector setup in the .geom file (loaded with the -g flag). +You can set the mininum and maximum 1/d values, in units of inverse meters, +with the options --min= and --max=. The default behaviour is to use the +entire detector extent from the geometry description file,loaded with the -g flag). -You can also adjust the number of histogram bins with: ---bins= +You can also adjust the number of histogram bins with the option --bins=, +where n is an integer. -Scaling can be set to produce linearly, quadratically or cubicly spaced 1/d -values using: --spacing= possible types are: +Scaling can be set to produce linearly, quadratically or cubically spaced 1/d +values using --spacing=. Possible types of spacing are: linear : linear (default) wilson : even spacing in Wilson plots (quadratic) volume : constant volume (cubic) .SH DATA TYPE OPTIONS -Powder_plot can accept .stream, .hkl, or HDF5 files as input parameters. For -the stream many choices exist for assembling the powder these can be accessed -using the --data= flag. Possible flags are: - reflection : uses peak positions from indexed reflection - hkl : uses the hkl list from indexed reflections (requires pdb - file) - d : uses the 1/d list from indexed reflections (default) - peaks : uses all peaks found from peak search - h5 : all points in HDF5, excluding bad regions - -Reflection differs from hkl or d as each crystal may index with a slightly -different unit cell size. The peaks option uses the peak lists and includes -and peaks (regardless if they indexed). - -A few other implemented flags to give better control are: - --no-sat-corr Don't correct values of saturated peaks using a table included in the HDF5 file. - --only-indexed Use with -data=peaks or h5 if you want to use the peak list of only indexed patterns. (Useful for seeing changes from patterns that do not index.) - --no-d-scaling Use with .hkl files if you want to not scale the powder by d^2 - --ring-corr Use if you want to scale the powder plot to correct for the fractional area sampled of the powder ring. (Useful for detectors with gaps.) - --use-redundancy Use with .hkl files if you want to use the number of measurements and not the number of symmetrical equivalent reflections as the number of time a reflection occurs in the powder - -.SH NOTE ON SUMMING H5 FILES - -Powder_plot does not sum a stack of HDF5 files from a list of HDF5 files. This -task is already accomplished with sum_stack. The output HDF5 file of sum_stack -can then be used as the input to powder_plot. This will produce the goal of -summing a list of HDF5 files into a powder. Also powder_plot can sum HDF5 files -if they are located in a stream file. This is useful with the --only-indexed, -to compare the powder from all the data to the power from only indexed patterns. +When taking input from stream, the d-spacing for a particular intensity can be +generated in a variety of different ways. You can choose which one with +--data=. Possible types are: + + reflection : use peak positions from indexed reflections + hkl : use the Miller indicies from indexed reflections, combined + with a unit cell from PDB file provided with -p. + d : use the 1/d values in the stream for indexed reflections, + which were calculated at the time of indexing using the + unit cell for the individual crystal. The individual + unit cell should be close to the actual cell, meaning + that this option should give a result close to "hkl". + peaks : use peak positions from peak search + h5 : use all pixels in the HDF5 file, excluding bad regions + +The default is "--data=d". + +.SH OPTIONS FOR MORE CONTROL + + --no-sat-corr +Don't correct values of saturated peaks using the table included in the HDF5 file. +See the help for indexamajig for more information. + + --only-indexed +Use with --data=peaks or h5 if you want to use the peak list of only indexed patterns. +This is useful for finding differences between patterns which could be indexed and +those which could not. + + --no-d-scaling +Do not scale the intensities in the powder plot by d^2. This should be used when +creating a powder plot from a reflection list. + + --ring-corr +Do not correct for the fractional area sampled of the powder ring. This might be +useful for detectors with gaps. + + --use-redundancy +Divide intensities by the number of measurements (the redundancy column in the +reflection list), and not the number of symmetrical equivalent reflections as the +number of times a reflection occurs in the powder. -- cgit v1.2.3