• Class: FleurinpData

  • String to pass to the DataFactory(): fleur.fleurinp

  • Aim: store input files for the FLEUR code and provide user-friendly editing.

  • What is stored in the database: the filenames, a parsed inp.xml files as nested dictionary

  • What is stored in the file repository: inp.xml file and other optional files.

  • Additional functionality: Provide user-friendly methods. Connected to structure and Kpoints AiiDA data structures



FleurinpData is an additional AiiDA data structure which represents everything a FleurCalculation needs, which is mainly a complete inp.xml file.


Currently, FleurinpData methods support ONLY inp.xml files, which have everything in them (kpoints, energy parameters, …), i.e which were created with the -explicit inpgen command line switch. In general it was designed to account for several separate files too, but this is no the default way Fleur should be used with AiiDA.

FleurinpData was implemented to make the plugin more user-friendly, hide complexity and ensure the connection to AiiDA data structures (StructureData, KpointsData). More detailed information about the methods can be found below and in the module code documentation.


If you want to change the input file use the FleurinpModifier (FleurinpModifier) class, because a FleurinpData object has to be stored in the database and usually sealed.


from aiida_fleur.data.fleurinp import FleurinpData
# or FleurinpData = DataFactory('fleur.fleurinp')

F = FleurinpData(files=['path_to_inp.xml_file', <other files>])
# or
F = FleurinpData(files=['inp.xml', <other files>], node=<folder_data_pk>)

If the node attribute is specified, AiiDA will try to get files from the FolderData corresponding to the node. If not, it tries to find an inp.xml file using an absolute path path_to_inp.xml_file.

Be aware that the inp.xml file name has to be named ‘inp.xml’, i.e. no file names are changed because the filenames will not be changed before submitting a Fleur Calculation. If you add another inp.xml file the first one will be overwritten.


  • inp_dict: Returns the inp_dict (the representation of the inp.xml file) as it will or is stored in the database.

  • files: Returns a list of files, which were added to FleurinpData. Note that all of these files will be copied to the folder where FLEUR will be run.

  • inp_version: Returns the version of the stored inp.xml

  • parser_info: Returns errors, warnings and information encountered while constructing the inp_dict from the inp.xml


FleurinpData will use the masci-tools library to parse the inp.xml. This library contains the schema files for the fleur input and output XML files for many of the fleur releases starting from version 0.27. If a version is encountered that is not yet stored in the installed version of the masci-tools library, the latest available version is used.

User Methods

Setting up atom labels

Label is a string that marks a certain atom in the inp.xml file. Labels are created automatically by the inpgen, however, in some cases it is helpful to control atom labeling. This can be done by setting up the kind name while initialising the structure:

structure = StructureData(cell=cell)
structure.append_atom(position=(0.0, 0.0, -z), symbols='Fe', name='Fe123')
structure.append_atom(position=(x, y, 0.0), symbols='Pt')
structure.append_atom(position=(0., 0., z), symbols='Pt')

in this case both of the Pr atoms will get default labels, but ‘Fe’ atom will the label ‘123’ (actually ‘ 123’, but all of the methods in AiiDA-Fleur are implemented in a way that user should know only last digit characters).


Kind name, which is used for labeling, must begin from the element name and end up with a number. It is very important that the first digit of the number is smaller than 4: 152, 3, 21 can be good choices, when 492, 66, 91 are forbidden.


Except setting up the label, providing a kind name also creates a new specie. This is because the 123 will not only appear as a label, but also in the atom number. In our case, the line in the inpgen input corresponding to Fe atom will look like 26.123 0 0 -z 123. Hence, if we would have another Fe atom with the default kind name, both of the Fe atom would belong to different atom group, generally resulting in lower symmetry.

Given labels can be used for simplified xml methods. For example, when one performs SOC calculations it might be needed to vary socscale parameter for a certain atom. Knowing the correct label of the atom, it is straightforward to make such a change in FleurinpData object (using the FleurinpModifier) or pass a corresponding line to inpxml_changes of workchain parameters:

# an example of inpxml_changes list, that sets socscale of the iron atom
# from the above structure to zero
inpxml_changes = [('set_species_label', {'at_label': '123',
                                        'attributedict': {
                                            'special': {'socscale': 0.0}
                                        'create': True

# in this example the atomgroup, to which the atom with label '222' belongs,
# will be modified
fm = FleurinpModifier(SomeFleurinp)
fm.set_atomgroup_label({'force': {'relaxXYZ': 'FFF'}, atom_label='                 222')