Differences

This shows you the differences between two versions of the page.

--- software:imaged11 [2018/07/13 14:37] – matthias
+++ software:imaged11 [2019/11/28 14:35] (current) – smerkel
@@ Line 8: / Line 8: @@
 First of all, you will need Linux, either as a separate harddrive (with lots of memory space) or as a virtual machine on your Windows/Apple computer. In case you are not familiar with Linux, ask your administrator for help and maybe do an online crash course to learn basic navigation and so on.
+Next things you need is //Anaconda// and //PIP//.
 ==== Installation ====
-Go to the [[https://github.com/FABLE-3DXRD/ImageD11|ImageD11 homepage]] and click on the green button "clone or download". Click on "Download ZIP". After the download is finished, navigate to the "Downloads" folder and unpack the ZIP folder to a directory of your choice.
+Go to the [[https://github.com/FABLE-3DXRD/ImageD11|ImageD11 homepage]] and click on the green button "clone or download". Click on "Download ZIP". The latest version is from January 2019. After the download is finished, navigate to the "Downloads" folder and unpack the ZIP folder to your python site-packages folder (e.g. /usr/lib64/python2.7/site-packages/). You might need to do it via the command line and with root access:
+    sudo unzip ImageD11-master.zip -d destination_folder
+The new ImageD11 version (January 2019) includes a file called //howto_compile_win64.txt//. If you follow its orders you shouldn't run into any problems during the installation. It includes instructions for your command line. Most of them are only necessary for a new computer but some of them are essential!
-Open the Konsole and navigate to the unpacked ImageD11 directory. There should be a folder called "setup.py". Type the following to the Konsole:
-    python setup.py build
-and after that:
-    python setup.py install
-If you get an error, there might be some of the additional libraries still missing. Check out the pages for the other software for more information. For example, Fabian requires many prerequisites. If the installation of ImageD11 doesn't seem to work, maybe try out another program first.
 ===== Starting the software =====
@@ Line 44: / Line 42: @@
 Try to start the program again (see above).
-==== Basic instructions ====
+===== Basic instructions =====
-The program should open. In //ImageD11// go to //Transformation//, then //Load filtered peaks// and choose the .flt file you would like to see (You automatically should be in the right folder.).
+The program usually opens as a long tool bar. Use your mouse to extend the window to a square shaped window. In the tool bar go to ''Transformation'', then ''Load filtered peaks'' and choose the //.flt// file you would like to see. You should remember which file you opened because this is not shown anywhere. Don't worry if nothing seems to happen when opening something.
-Then go to //Transformation// and //plot y/z//. You should see a 2D diffraction image of all the spots which passed the threshold. From now on, we call this image //diffraction plot//. You should remember, which file you opened, because ImageD11 doesn't show this anywhere.
+==== Visualization in 2D ====
+There are two basic plotting options in //ImageD11//:
+  * The standard view of a diffraction image
+  * The cake view
+Note that //ImageD11// can show only one plot at once. If you want to compare two (or more) plots, you can either save the plots (as .png file) or open a new //ImageD11// window in a new tab of the command line.
-//ImageD11// can show only one //diffraction plot// at once. If you want to compare two (or more) //diffraction plots//, you can either save the plots (as .png file) or open a new //ImageD11// in a new tab of the Konsole.
+To switch between the two plot types, you can simply erase the current plot with ''Clear'' and plot the other one. Note that ''Clear'' only erases the plot, not the data!
+You can zoom in the plots with your mouse (left click). Unzoom with right click.
+=== 2D diffraction image ===
+Click ''Transformation'' and ''plot y/z''. You should see a 2D diffraction image of all the spots which passed the threshold.
+=== Cake image ===
+Click ''Transformation'' and ''plot tth/eta''. The cake view of all spots which passed the threshold appears.
 ==== Editing parameters ====
+=== Entering parameters ===
 {{ software:imaged11_parameters.png?300|Screenshot of the parameter window}}
-After you loaded a file you should modify the parameters. For this, click on //Transformation// --> //Edit parameters//. You will see a window like this one. Enter all necessary parameters. Necessary are usually these ones:
+After you loaded a file you should modify the parameters. For this, click on ''Transformation'' and ''Edit parameters''. You will see a window like the one to the right. Enter all necessary parameters. Necessary are usually these ones:
-  * Cell parameters: cell_a, cell_b, cell_c in Angstrom; cell_alpha, cell_beta, cell_gamma in degrees
+  * **Cell parameters**: cell_a, cell_b, cell_c in Angstrom; cell_alpha, cell_beta, cell_gamma in degrees (typically taken from [[software:maud|Maud]] refinement)
-  * Lattice type: P = primitive, F = face-centered, A = A-centered, ...
+  * **Lattice type**: P = primitive, F = face-centered, A = A-centered, ...
-  * Distance = sample-detector distance in µm
+  * **Distance** = sample-detector distance in µm (typically taken from [[software:maud|Maud]] refinement)
-  * O-matrix (same as in the //.inp// file)
+  * **O-matrix** (check the [[dac_experiments:geometry|concept of the O-matrix]] if you are not sure)
-  * Omegasign = rotation direction (1 = clockwise, -1 = counterclockwise)
+  * **Omegasign** = rotation direction (1 = clockwise, -1 = counterclockwise)
-  * Tilt (only if you know there is tilt)
+  * **Tilt**: Be careful with this one: Tilt in ImageD11 is sometimes not the same as tilt in other software! (usually keep 0 and refine later)
-  * Wavelength in Angstrom
+  * **Wavelength** in Angstrom
-  * y_size, z_size: size of one pixel in y direction and z direction in µm
+  * **y_size**, **z_size**: size of one pixel in y direction and z direction in µm
-  * y_center, z_center: pixel coordinates of the beam center (if your detector has 2048x2048 pixels the center is usually somewhere close to 1024x1024)
+  * **y_center**, **z_center**: pixel coordinates of the beam center (usually taken from [[software:dioptas|Dioptas calibration]]: ''y_center = Dioptas Center X'' and ''z_center = 2048 - Dioptas Center Y'')
+  * Supplementary parameter **cell_sg**. You can also provide your material space group. This has to be done manually with a text editor and can not be done using the Graphical User Interface. If the space group is provided, ImageD11 will use xfab for generate unique hkls.
+=== Typical mistakes during parameter entering ===
+  * You entered a value in the wrong order of magnitude, for example: ''detector distance'' = ''400'' (instead of ''400000'' because this value has to be entered in µm, not in mm)
+  * You entered a value in the wrong unit, for example: ''y_size'' = ''1024'' (instead of ''200'', because the software needs the individual pixel size of 200 µm, not the total amount of pixels in y direction)
+=== Saving parameters ===
+You can also store parameter lists with ''Transformation'' and ''Save parameters''. Note that during storage you sometimes have to add the extension //.par// manually. To recall the list, click ''Transformation'' and ''Load parameters''.
+=== Refinement of parameters ===
+The values you enter by hand are usually not 100% correct. This is because //ImageD11// calculates or labels certain parameters different than other software does.
-The values you enter are usually not 100% correct. When you plot the //tth/eta plot// you usually won't see totally straight vertical lines, especially not when you zoom in. So ImageD11 has an option for refinement implemented. To refine a parameter, you have to activate the //Vary?// box to the right of the parameter (of course, you can refine several parameters at once). Afterwards, close the window with //Ok// and go to //Transformations// --> //Fit//. The spots will move. Repeat clicking on //Fit// until most of the vertical lines are totally straight.
+When you plot the cake view (''tth/eta plot'') you usually won't see totally straight vertical lines, especially not when you zoom in. So ImageD11 has an option for refinement implemented. To refine a parameter, you have to activate the ''Vary?'' box to the right of the parameter (of course, you can refine several parameters at once). Afterwards, close the window with ''Ok'' and go to ''Transformation'' and ''Fit''. The spots will move a bit. Repeat clicking on ''Fit'' until most of the vertical lines are totally straight.
-Some advice for your refinement
+**Some advice for your refinement**
-  * If your spots sit on a sinus-shaped line of exactly one period, refine only y_center and z_center. This should already solve the problem.
+  * Don't refine when you have a zoomed view. Always refine when you are fully unzoomed.
+  * If your spots sit on sinus-shaped lines of exactly one period, refine only y_center and z_center. This should already solve the problem.
   * Don't refine too many parameters at once!
   * Don't refine parameters you know for sure, such as the angle of the unit cell or the wavelength! (Especially don't refine strain, if this is a simulation without strain!)
+  * When you see that your refinement gets worse and worse, you should enter your starting values again and start your refinement from the beginning.
-Typical mistakes during parameter entering and refinement
-  * You entered a value in the wrong order of magnitude, for example: //detector distance = 400// (instead of 400000 because you need to enter this value has to be entered in µm, not in mm)
-  * You entered a value in the wrong unit, for example: //y_size = 1024// (instead of 200, because the software needs the individual pixel size of 200 µm, not the total amount of pixels in y direction)
-  * You refine too many parameters at once