Top level (root)¶
This node represents the top level of the HDF5 file and holds some general information about the file.
- A colon separated list that shows which components are present in the file. The only mandatory component is exchange. A more general Data Exchange file also contains measurement and process information, if so these will be declared in implements as exchange:measurement:process
- exchange or exchange_N
- The data taken from measurements or processing. Dimension descriptors within the group may also serve to describe “positioner” values involved in a scan.
- measurement or measurement_N
- Description of the sample and instrument as configured for the measurement. This group is appropriate for relatively static metadata. For measurements where there are many “positioner” values (aka a “scan”), it is more sensible to add dimension(s) to the exchange dataset, and describe the “positioner” values as dimension scales rather than record the data via multiple matching measurement and exchange groups. This is a judgement left to the user.
- The Process group describes all the “work” that has been done. This includes data processing steps that have been applied to the data as well as experimental steps (e.g. data collection strategy etc.) and sample preparation ahead of the experiment and during the measurement (e.g. environment conditions etc.).
The exchange group is where scientific datasets reside. This group contains one or more array datasets containing n-dimensional data and optional descriptions of the axes (dimension scale datasets). Exactly how this group is used is dependent on the application, however the general idea is that one exchange group contains one cohesive dataset. If, for example, the dataset is processed into some other form, then another exchange group is used to store the derived data.
Multiple exchange groups are numbered consecutively as exchange_N. At a minimum, each exchange group should have a primary dataset named data. The title is optional.
|description||string dataset||“raw absorption tomo”|
|data||array dataset||n-dimensional dataset|
Table: Exchange Group Members
- Descriptive name for data dataset. Current types include: absorption_tomography, phase_tomography, dpc_tomography
- The primary scientific dataset. Additional related datasets may have any arbitrary name. Each dataset should have a units and description attribute. Discussion of dimension descriptors and optional axes attribute is covered in Section [sec:multidims].
Description and units can be added as attribute to any data, both array or values, inside a data exchange file. If units is omitted default is SI.
Table: data attributes
This group holds sample and instrument information. These groups are designed to hold relatively static data about the sample and instrument configuration at the time of the measurement. Rapidly changing positioner values (aka scan) are better represented in the exchange group dataset.
Table: Measurement Group Members
- The instrument used to collect this data.
- The sample measured.
The instrument group stores all relevant beamline components status at the beginning of a measurement. While all these fields are optional, if you do intend to include them they should appear within this parentage of groups.
- Name of the instrument.
- List of components part of the instrument. Replace component with the actual item name, source, mirror, etc.
- The detectors that compose the instrument.
Class describing the component being used.
|arbitrary_label_1||string dataset||“what ever”|
|arbitrary_label_2||string dataset||“what ever”|
|arbitrary_label_n||string dataset||“what ever”|
Table: Component Description
- Date and time source was measured.
Logging instrument and beamline component setup parameters (static setup values) is not defined by Data Exchange because is specific and different for each instrument and beamline. To capture this information Data Exchange requires to set a setup group under each beamline component and leaves each facility free to store what is relevant for each component (list of motor positions etc.). Ideally each component in the instrument list (source, shutter, attenuator etc.) should have included its setup group. For setup values not associated with a specific beamline component a setup group in the instrument group should be created.
Table: Setup Group Members
The geometry group is common to many of the subgroups under measurement. The intent is to describe the translation and rotation (orientation) of the sample or instrument component relative to some coordinate system. Since we believe it is not possible to determine all possible uses at this time, we leave the precise definition of geometry up to the technique. We do encourage the use of separate translation and orientation subgroups within geometry. As such, we do not describe geometry further here. This class holds the general position and orientation of a component.
- The position of the object with respect to the origin of your coordinate system.
- The rotation of the object with respect to your coordinate system.
This is the description for the general spatial location of a component for tomography.
|distances||3 float array dataset||(0, 0.001, 0)|
- The x, y and z components of the translation of the origin of the objectrelative to the origin of the global coordinate system (the place wherethe X-ray beam meets the sample when the sample is first aligned in the beam).If distances does not have the attribute units set then the units are inmeters.
This is the description for the orientation of a component for tomography.
|value||6 float array dataset|
- Dot products between the local and the global unit vectors. Unitless
The orientation information is stored as direction cosines. The direction cosines will be between the local coordinate directions and the global coordinate directions. The unit vectors in both the local and global coordinates are right-handed and orthonormal.
Calling the local unit vectors (x’, y’,z’) and the reference unit vectors (x, y, z) the six numbers will be
is the scalar dot product (cosine of the angle between the unit vectors).
Notice that this corresponds to the first two rows of the rotation matrix that transforms from the global orientation to the local orientation. The third row can be recovered by using the fact that the basis vectors are orthonormal.
This group holds basic information about the sample, its geometry, properties, the sample owner (user) and sample proposal information. While all these fields are optional, if you do intend to include them they should appear within this parentage of groups.
|name||string dataset||“cells sample 1”|
|description||string dataset||“malaria cells”|
|preparation_date||string dataset (ISO 8601)||“2012-07-31T21:15:22+0600”|
|chemical_formula||string dataset (abbr. CIF format)||“(Cd 2+)3, 2(H2 O)”|
|position||string dataset||“2D” APS robot coord.|
Table: Sample Group Members
- Descriptive name of the sample.
- Description of the sample.
- Date and time the sample was prepared.
- Sample chemical formula using the CIF format.
- Mass of the sample.
- Sample environment.
- Sample temperature.
- Sample temperature set point.
- Sample pressure.
- Sample thickness.
- Sample position in the sample changer/robot.
- Sample center of mass position and orientation.
- Facility experiment identifiers.
- Experimenter identifiers.
This provides references to facility ids for the proposal, scheduled activity, and safety form.
Table: Experiment Group Members
- Proposal reference number. For the APS this is the General UserProposal number.
- Proposal scheduler id. For the APS this is the beamline scheduler activity id.
- Safety reference document. For the APS this is the ExperimentSafety Approval Form number.
Description of a single experimenter. Multiple experimenters can be represented through numbered entries such as experimenter_1, experimenter_2.
|name||string dataset||“John Doe”|
|role||string dataset||“Project PI”|
|affiliation||string dataset||“University of California, Berkeley”|
|address||string dataset||“EPS UC Berkeley CA 94720 4767 USA”|
|phone||string dataset||“+1 123 456 0000”|
Table: Experimenter Group Members
name: User name.
role: User role.
affiliation: User affiliation.
address: User address.
phoen: User phone number.
email: User e-mail address
facility_user_id: User badge number
Process is the documentation of the data collection strategy (acquisition) steps, all transformations, analyses and interpretations of data performed by a sequence of process functions (actor) as well as any sample preparation step done ahead of the experiment and during the measurement (e.g. environment conditions etc.).
Maintaining this history, also called provenance, allows for reproducible data. The Data Exchange format tracks process by allowing each actor to append process information to a process table.
The process table tracks provenance in the execution order as a series of processing steps by appending sequential actor entries in the process table.
Table: Process Group Members
- Descriptive process task.
- Description of the process task.
This is the actor description group. Each entry of the process table will refer to the correspondent actor description.
|name||string dataset||“test rec”|
Table: Actor Group Members
- Descriptive actor task.
- Description of the actor task.
- Version of the actor task.If available this can be the repository link to the actor version used
- input_data, output_data
- Origin and destination of the data processed by the actor.
Here is where to log the actor setup parameters (static setup values).
Table: Actor Setup Group
Scientific users will not generally be expected to maintain data in this group. The expectation is that the data collection and analysis pipeline tools will automatically record process steps using this group. In addition, it is possible to re-run an analysis using the information provided here.
|actor_1||21:15:22||21:15:23||SUCCESS||OK||/process/actor_1||raw data collection|
|actor_n||21:17:28||22:15:22||QUEUED||OK||/process/actor_n||transfer data to user|
Table: Process table to log actors activity
- Name of the process in the pipeline stage that is executed at this step.
- Time the process started.
- TIme the process ended.
- Current process status. May be one of the following: QUEUED,RUNNING, FAILED, or SUCCESS.
- A process specific message generated by the process. It may be aconfirmation that the process was successful, or a detailed errormessage, for example.
- Path to the actor description group. The process description groupcontains all metadata to perform the specific process. Thisreference is simply the HDF5 path within this file of thetechnique specific process description group. The processdescription group should contain all parameters necessary to runthe process, including the name and version of any externalanalysis tool used to process the data. It should also containinput and output references that point to theexchange_N groups that contain the input and outputdatasets of the process.
- Process description.