Table of Contents
These tutorials assume you have read, understood and completed the steps detailed at the following links:
For these examples we will be using an arbitrary instrument, and any differences between other instruments will be clearly documented.
These tutorials will use the data found in the ISIS example data sets. The data sets can be downloaded from the Mantid download page, under Sample datasets - ISIS. Users may also use their own data instead however results will differ. Some additional configuration may also be required which is explained in later tutorials.
This tutorial will help you setup the ISIS Powder diffraction scripts for your instrument. It assumes no prior knowledge or usage of the scripts and a fresh install of Mantid.
Open your Mantid install location, by default this will be C:\MantidInstall on Windows and /opt/Mantid on Linux. Open scripts, diffraction, isis_powder. In these tutorials we will be using Polaris examples and data however you may set-up a different instrument.
Open polaris_routines (or ‘instName’_routines), there will be a folder called examples. Copy the contents (all files and folders) within the examples folder to a known location.
Open up Mantid, then go to the scripting window by either pressing F3 or going to View - Script Window
First we need to import the instrument to be able to use it. If you are using a different instrument substitute Polaris for your instrument name:
# Note this IS case sensitive
from isis_powder import Polaris
This section is for script writers who are not familiar or comfortable with object orientation. If you feel comfortable with the concept of objects holding state you may skip onto the next section.
The ISIS Powder scripts use objects to hold a state, an object is best illustrated with the below example:
blue_duck = Duck(type="Blue")
red_duck = Duck(type="Rubber duck")
On lines 1 and 2 we create a new duck object. Each object has a name we choose (in this case blue_duck and red_duck) and a separate state, but the same actions we can perform on each are the same. For example
blue_duck.feed()
We now have fed blue_duck so its state will have changed so it is no longer hungry. However the red_duck has not changed at all so its state is still hungry in this example.
Because objects have their own state you can create multiple objects in your script to perform different actions, such as processing half your data with one set of options and the other half of the data with another set of options.
Navigate back to the files copied from this section of the tutorial Copying instrument example files. There should be two files and a folder. If you are using the ISIS example data set (see Obtaining tutorial data) you will not need to modify anything at this point.
If you are not using the ISIS example data set you will need to modify your calibration directory and cycle mapping as detailed here: Cycle mapping files
Take notes of the following paths as we will need them later:
For example in the POLARIS example folder these filenames will be:
Having introduced objects in: A quick introduction to objects we can now go ahead and create an instrument object.
from isis_powder import Polaris
# This name is arbitrary
a_pol_obj = Polaris()
If you try to run this code the script will complain whenever it comes across a parameter it requires but has not been set. The following parameters must be set for all instruments:
There will also be additional instrument specific parameters required, a list of these can be found in the relevant instrument reference: Instrument Reference for example all instruments require a cycle mapping file. On GEM and POLARIS this is called the calibration_mapping_file, on PEARL this is the calibration_config_path.
Using the above information we can start to populate the required parameters (see Paths to the required files for where these paths came from):
from isis_powder import Polaris
a_pol_obj = Polaris(user_name="Your name here",
calibration_directory=*Path to calibration directory*,
calibration_config_path=*Path to folder*\\*cycle mapping name.YAML*,
....etc.)
Each time we execute the code it will inform us if a parameter is required at that point and we have forgotten to enter it. When you see script execution finished it means we have enough information to create the instrument object.
In the next tutorial we will focus a vanadium run and use that to focus a standard sample.
This tutorial assumes you have followed the steps in the previous tutorial Setting up and have created an instrument object successfully.
We now have an object for the instrument we specified, if you followed the previous tutorial this will be a Polaris object. These objects have methods we can access using their ‘.’ operator. We will use this to create a vanadium run on Polaris:
1 2 3 4 | from isis_powder import Polaris
a_pol_obj = Polaris(...)
a_pol_obj.create_vanadium(...)
|
On line 4 we call the create_vanadium method on the Polaris object, all instruments will have this method however the parameters they accept and require are bespoke. Parameters can be found for each individual instrument in the reference document: Instrument Reference
Additionally as the objects hold state we can set a parameter anywhere. For example on Polaris the mode indicates the chopper state for this/these run(s). This can either be set when we create the object like this:
from isis_powder import Polaris
a_pol_obj = Polaris(mode="PDF", ....)
a_pol_obj.create_vanadium(...)
Or set whilst calling a method like this:
from isis_powder import Polaris
a_pol_obj = Polaris(...)
a_pol_obj.create_vanadium(mode="PDF", ...)
Both of the above are equivalent. Additionally if we change the value the scripts will warn us. This can be demonstrated with the following example:
from isis_powder import Polaris
a_pol_obj = Polaris(mode="PDF", ...)
# The following line will warn us we changed the chopper
# status from PDF to Rietveld. It will also remain
# in Rietveld mode from now on till we change it again
a_pol_obj.create_vanadium(mode="Rietveld", ...)
# Mode is still Rietveld on the following line
a_pol_obj.create_vanadium(...)
For these reasons it is recommended to create multiple objects when you need to switch between different settings within a script:
from isis_powder import Polaris
pol_PDF = Polaris(mode="PDF", ...)
pol_Rietveld = Polaris(mode="Rietveld", ...)
# Runs with the chopper set to PDF mode:
pol_PDF.create_vanadium(...)
# Runs with the chopper set to Rietveld mode:
pol_Rietveld.create_vanadium(...)
Because of the way objects hold state in ISIS Powder (see: How objects hold state in ISIS Powder) it is up to the reader of this tutorial where they set different parameters.
As previously mentioned each instrument has bespoke parameters and can be found in the individual instrument reference document: Instrument Reference
Additionally as noted previously this tutorial assumes the user is using the example ISIS data set ( see: Obtaining tutorial data). If they are not they will need to setup their cycle mapping to their data detailed here: Cycle mapping files
For Polaris we require the following parameters in addition to the parameters discussed to create the object (see Creating the instrument object):
Note: Due to the complexity of the Polaris instrument definition it will take Mantid up to 10 minutes to load your first data set for this instrument.
As we will be later focusing run number 98533 we can use that to ensure the correct cycle is selected for the first_cycle_run_no input.
from isis_powder import Polaris
# This should be set from the previous tutorial.
a_pol_obj = Polaris(....)
a_pol_obj.create_vanadium(first_cycle_run_no=98533,
do_absorb_corrections=True,
mode="Rietveld",
multiple_scattering=False)
Executing the above should now successfully process the vanadium run, you should have two resulting workspaces for the vanadium run in dSpacing and TOF. Additionally there will be another workspace containing the splines which will be used when focusing future data.
Having successfully processed a vanadium run (see: Creating the vanadium run) we are now able to focus a data set. For this tutorial we will be focusing a sample of Silicon.
It is highly recommended to create a separate script file for focusing data, this ensures the vanadium is not reprocessed every time data is focused.
To focus data we can call the focus method present on all instruments. As previously mentioned each instrument has bespoke parameters, these can be found in the individual instrument reference document: Instrument Reference
from isis_powder import Polaris
# This should be set from the previous tutorial.
a_pol_obj = Polaris(....)
a_pol_obj.focus(...)
To focus the Si sample included in the ISIS data set we require the following parameters:
For this tutorial the run number will be 98533, and input_mode will not affect the result as it is a single run. Additionally in the example data you could focus 98534 (YAG sample) too.
from isis_powder import Polaris
# This should be set from the previous tutorial.
a_pol_obj = Polaris(....)
a_pol_obj.focus(input_mode="Individual", run_number=98533,
mode="Rietveld",
do_absorb_corrections=False,
do_van_normalisation=True)
This will now process the data and produce two workspace groups for the results in dSpacing and TOF in addition to another group containing the spline(s) used whilst processing the data.
Congratulations you have now focused a data set using ISIS Powder.
After focusing the data it is saved in a variety of formats which suits the instrument. These can be found in the user specified output directory. The scripts will automatically create the label for the current cycle (covered in additional detail later Cycle mapping files).
Within the label folder a new folder will be created or used matching the user_name specified. Within that folder will be the output data in the various formats that is used on that instrument to perform data analysis.
This tutorial assumes you have successfully created an instrument object as described here: Creating the instrument object.
You have probably noticed that a lot of the parameters set do not change whenever you create an instrument object and a warning is emitted stating you are not using a configuration file.
The rational behind a configuration file is to move settings which rarely change but are machine specific to a separate file you can load in instead. For example the output directory or calibration directory do not change often.
Navigate back to the files copied from the example folder (see: Copying instrument example files). There is a file we have not been using which will be named along the lines of ‘inst’_config_example.YAML.
This will come pre-configured with some examples of how parameters are set in the files. The names always match parameter names which can be found in the instrument reference documentation: Instrument Reference
For example if we currently have the output directory as follows:
from isis_powder import Polaris
# Note the r before " avoids us having to put \\
a_pol_obj = Polaris(output_directory=r"C:\path\to\your\output_dir", ....)
We can instead move it to the YAML file so it reads as follows:
# YAML FILE:
# Note the single quotes on a path in a YAML file
output_directory: 'C:\path\to\your\output_dir'
Additionally we can move parameters which should be defaults into the same file too:
#YAML FILE:
output_directory: 'C:\path\to\your\output_dir'
do_van_normalisation: True
Warning
Within the YAML files the most recent value also takes precedence. So if user_name appeared twice the value closest to the bottom will be used. This is implementation specific and should not be relied on. Users should strive to ensure each key - value pair appears once to avoid confusion.
You will need to make a note of the full path to the configuration file. Note that the filename entered must end with .YAML (even if it is not shown when browsing the files on your OS).
Setting the configuration file from the previous example we now have a default output directory and perform vanadium normalisation by default too.
from isis_powder import Polaris
config_file_path = r"C:\path\to\your\config_file.YAML"
a_pol_obj = Polaris(config_file=config_file_path, ...)
# Will now divide by the vanadium run by default as this was
# set in the configuration file
a_pol_obj.focus(...)
Any property set in the configuration file can be overridden. So if you require a different output directory for a specific script you can still use the original configuration file.
from isis_powder import Polaris
config_file_path = r"C:\path\to\your\config_file.YAML"
# Output directory changed to our own output directory,
# and warning emitted informing us this has happened
a_pol_obj = Polaris(config_file=config_file_path,
output_dir=r"C:\path\to\new\output_dir", ...)
# As the object has a state it will still be set to our custom
# output directory here (instead of configuration one) without
# restating it
a_pol_obj.focus(...)
It is recommended instrument scientists move optimal defaults (such as performing vanadium normalisation) into a configuration file which the scripts use.
The cycle mapping file is used to hold various details about the current and past cycles. These details include the empty and vanadium run number(s), current label and offset filename.
The label is used to separate output data into its various cycle numbers, Mantid will correctly handle the cycle on input data. The goal of the label is to ensure runs end up in the output folder the user wants them in, regardless of which cycle ISIS is on.
These examples explain the layout and concept of YAML files. For instrument specific examples please look at the individual instrument reference document: Instrument Reference for an example specific to your instrument.
The simplest example of the calibration file is used on Pearl as the empty, label and vanadium are the same regardless of mode.
# This is the layout used on PEARL
# NB this example is not representative of actual run numbers
123-200:
# Notice how the indentation changes to indicate it belongs
# to this section
label : "1_2"
vanadium_run_numbers : "150"
empty_run_numbers : "160"
offset_file_name : "pearl_offset_1_2.cal"
On GEM the two chopper modes PDF and Rietveld affect the empty and vanadium run numbers used. In this case the additional indentation underneath the respective mode is used.
Fields can be left blank until a later date if runs in different modes have not been collected yet.
# This is the layout used on GEM
# NB this example is not representative of actual run numbers
123-200:
label: "1_2"
offset_file_name: "offsets.cal"
PDF:
# Blank entries are allowed provided we do not try to run in PDF mode
vanadium_run_numbers: ""
empty_run_numbers: ""
# Notice it is not case sensitive
rietveld:
# The indentation indicates these are for Rietveld mode
vanadium_run_numbers: "130"
empty_run_numbers: "131"
The run numbers for a cycle use the same syntax as the run number field. You can specify ranges of runs, have gaps or individual runs. For example 100-103, 105 will specify runs 100, 101, 102, 103 and 105.
The mapping also allows unbounded runs, this is useful for a cycle that is in progress as the final run number of a cycle is unknown
1-122:
label : "1_1"
...
123-:
label : "1_2"
...
All runs from 1-122 inclusive will go use the details associated with label 1_1, whilst any runs after 123 will use label 1_2. These values also have validation to ensure that there is only one unbounded range and no values come after the starting interval. For example in the above example adding a section for runs 200- or 200-210 would fail validation.
The user specified calibration directory directly relates to a cycle mapping file. After writing or adapting a cycle mapping file for your instrument you must update the calibration directory. Using the cycle mapping from Peal:
# NB this example is not representative of actual run numbers
123-200:
label : "1_2"
vanadium_run_numbers : "150"
empty_run_numbers : "160"
offset_file_name : "pearl_offset_1_2.cal"
The relevant fields from the cycle mapping are the label and offset_file_name. Within the calibration directory a folder with the label name must exist and contain a cal file with the offset_file_name.
In this example we need a folder within the calibration directory called 1_2 which holds a cal file called pearl_offset_1_2.cal.
The splines of the processed vanadium uses the run number and offset file name as a fingerprint to uniquely identify it. Because of this we can have two sets of details corresponding to the same cycle.
# NB this example is not representative of actual run numbers
123-150:
label : "1_2"
vanadium_run_numbers : "150"
empty_run_numbers : "152"
offset_file_name : "pearl_offset_1_2.cal"
151-200:
label : "1_2"
# Notice the changed details for runs 151 onwards
vanadium_run_numbers : "170"
empty_run_numbers : "160"
offset_file_name : "pearl_offset_1_2-second.cal"
The scripts also support processing intermediate files. This tutorial assumes you have successfully focused data previously as detailed here: Focusing a data set.
To process intermediate runs for example .s01 or .s02 files you must ensure the user directories are setup to include the folder where these files are located.
The instructions for this can be found here: Prerequisites. Note: The ‘Search Data Archive’ option will not locate intermediate runs as only completed runs are published to the data archive.
To indicate the extension to process the file_ext can be specified like so:
from isis_powder import Polaris
a_pol_obj = Polaris(....)
a_pol_obj.focus(file_ext="s01", ...)
# Or
a_pol_obj.focus(file_ext=".s01", ...)
This will locate a .s01 file for that run number and focus it like a normal run. The output filename will also reflect that this is a partial file. For run number 123 and file extension s1 the output filename will be s01’InstrumentName‘123.nxs. This allows users to easily distinguish between full runs and partial runs in the output folder. (For more details about the output folder see Output files)
This tutorial assumes you have successfully focused data previously as detailed here: Focusing a data set.
To perform absorption corrections on a sample we must first specify the chemical properties of the sample by creating a sample properties object. (See A quick introduction to objects.)
Note: Not all instruments support sample absorption corrections. Please check the instrument reference: :ref:`instrument_doc_links_isis-powder-diffraction-ref`. If the instrument has a *set_sample_details method it supports sample absorption corrections*
First we need to import the sample details object from ISIS Powder. The properties required when creating a SampleDetails object is the geometry of the sample.
Note: this assumes a cylinder geometry
For more details see SetSample v1.
from isis_powder import Polaris, SampleDetails
# Creates a cylinder of height 3.0, radius 2.0
# at position 0, 1, 2 (x, y, z)
position = [0, 1, 2]
# Create a new sample details object
my_sample = SampleDetails(height=3.0, radius=2.0, center=position)
Having set the sample geometry we can now set the chemical material and optionally the number density. If the chemical formula is not a single element the number density must be entered as it cannot be calculated.
For accepted syntax of chemical formulas see SetSampleMaterial v1. Specifically the section on specifying chemical composition if you are using isotopes. This will allow Mantid to automatically calculate the properties except for number density.
The material must be set before absorption corrections can be calculated for a sample.
... snip from previous example ...
my_sample = SampleDetails(height=3.0, radius=2.0, center=position)
my_sample.set_material(chemical_formula="V")
# OR
my_sample.set_material(chemical_formula="VNb", number_density=123)
Advanced material properties can be optionally set instead of letting Mantid calculate them. These properties are:
Note: This is purely optional and Mantid will calculate these values based on the chemical formula entered if this is not set
... snip from previous example ...
my_sample = SampleDetails(height=3.0, radius=2.0, center=position)
my_sample.set_material(chemical_formula="VNb", number_density=123)
# Setting individual properties:
my_sample.set_material_properties(absorption_cross_section=123,
scattering_cross_section=456)
Having created a new SampleDetails object (Create SampleDetails object) and then set the chemical material (Setting the material details) we can instruct the scripts to use these details whilst focusing.
This is done by calling set_sample_details on the instrument object, this will then use those sample details each time absorption corrections are applied to the sample. (See How objects hold state in ISIS Powder)
from isis_powder import Polaris, SampleDetails
... snip from previous examples ...
my_sample = SampleDetails(...)
my_sample.set_material(...)
polaris_obj = Polaris(...)
polaris_obj.set_sample_details(sample=my_sample)
# Indicate we want to perform sample absorption corrections whilst focusing
polaris_obj.focus(do_absorb_corrections=True, ...)
Warning
This method is not recommended for changing multiple samples. Instead it is recommended you create a new sample details object if you need to change properties mid way through a script. See Create SampleDetails object and A quick introduction to objects.
Note: The geometry of a sample cannot be changed without creating a new sample details object
Once you have set a material by calling set_material or set the properties by calling set_material_properties you will not be able to change (or set) these details without first resetting the object. This is to enforce the sample properties being set only once so that users are guaranteed of the state.
If you wish to change the chemical material or its advanced properties without creating a new sample details object you can call reset_sample_material. This will reset all details (i.e advanced properties and chemical properties)
from isis_powder import Polaris, SampleDetails
my_sample = SampleDetails(...)
my_sample.set_material(...)
# Next line will throw as it has already been set once
my_sample.set_material(...)
# This is still ok as its first time
my_sample.set_material_properties(...)
# Reset material
my_sample.reset_sample_material()
# Now allowed as object does not have a chemical formula associated
my_sample.set_material(...)
Warning
This section is intended for instrument scientists. The advanced configuration distributed with Mantid use optimal values for each instrument and should not be changed unless you understand what you are doing.
Note: Parameters should not be changed in the advanced configuration for a few runs. If you require a set of values to be changed for a range of runs (such as the cropping values) please set the value in the scripting window or configuration file instead (see: :ref:`configuration_files_isis-powder-diffraction-ref`).
The advanced configuration file provides optimal defaults for an instrument and applies to all runs unless otherwise specified. If this file is modified Mantid will not remove it on uninstall or reinstall, or upgrade. (Note: This behavior is not guaranteed and should not be relied on)
It is highly recommended you read the instrument reference found here: Instrument Reference to understand the purpose of each property and the effect changing it may have.
If you change any values in your advanced properties file could you please forward the new value to the Mantid development team to ensure this new value is distributed in future versions of Mantid
For the purposes of testing a parameter can be overridden at script runtime. The hierarchy of scripts is: scripting window > config file > advanced config. In other words a value set in the configuration file will override one found in the advanced configuration file. A value set in the scripting window will override one found in the configuration file.
A warning will always be emitted when a value is overridden so that the user is fully aware when this is happening.
For example to test a different spline coefficient value
from isis_powder import Polaris
a_pol_obj = Polaris(spline_coefficient=80, ...)
a_pol_obj.create_vanadium(...)
This will create a new vanadium run with the spline coefficient set to 80. Note that until create_vanadium is run again in this example any future data will implicitly use the splines with a coefficient of 80.
If you wish to change or view the advanced configuration files these can be found under MantidInstall/scripts/diffraction/isis_powder/inst _routines and will be called inst _advanced_config.py
If you change a value within the advanced config file you will need to restart Mantid for it to take effect. If you are happy with the new value please ensure you forward it on to the Mantid development team to be distributed in future versions.
Category: Techniques