It is possible to save data loaded in AlliGator (and reload them!) in a file format based on the HDF5 hierarchical format developed and maintained by the HDF Group.
This page briefly describes the file format used by AlliGator. Note that the latest AlliGator version supports all previous HDF5 file format versions.
Two new fields (MAC Address and Windows Username) have been added to the File Information group, while an additional SYNC Period parameter was added to the DAQ Parameters cluster.
The resulting structure is thus:
File Information group containing 9 strings, 2 integers and 1 Boolean:
DAQ Parameters cluster (saved as a Table): this is unchanged since v 0.1 and is reproduced here for convenience:
A new field (Compression) was added to the File Information group to specify whether compression was used (skipping compression speeds up data saving and loading). In principle, this should have no impact on HDF5 loading, as the HDF5 library is capable of determining this without user input. This field was added for information only.
File Information group containing 7 strings, 2 integers and 1 Boolean:
Additional information on the part of the sensor data saved in the file, a new Image Information group has been added as described below.
Image Information group contains 2 clusters:
Notes:
* Booleans are stored as integers in HDF5 files (FLASE: 0, TRUE: 1)
** Ux: x-bit unsigned integer
*** Ix: x-bit signed integer
v 0.3 (2021-01-29)
In order to support asynchronous data saving by third party software, version 0.3 changes the way the individual gate images are saved. Instead of being all clumped into a monolithic 3D array, each gate image is saved as a separate 2D array, as described below. This has the additional advantage that it simplifies accessing individual gates. As in v 0.2, each image gate is compressed, which significantly saves space, without incurring any significant decoding time cost.
The file is comprised of:
File Information group containing 7 strings and 2 integers:
DAQ Parameters cluster (saved as a Table): this is unchanged since v 0.1 and is reproduced here for convenience:
Gate Images group containing the gate images in the dataset as a series of 2D arrays labeled Gate n, where n will typically run from 1 to # Gates (the parameter found in the DAQ Parameters cluster described above.
Note: there is no requirement that the number of gate images stored in the file matches the number of gates provided by the # Gates parameter. For instance, if an acquisition is interrupted before the total number of gates has been acquired, the file will be closed without the full complement of gates.
There is also no check that the number of gate images remains smaller or equal to that parameter. User beware!
Metadata string contains any other information that does not fit in any of the previous fields. There is currently no specified formatting. It is likely that this will change in the future.
v 0.2
HDF5 files saved in this version are identical to v 0.1 files except that the Gate Images data is compressed using the deflate algorithm, with a level 9 compression and block size equal to 1x1xG where G is the number of gates.
The file is comprised of:
File Information group containing two strings:
Parameter Name
File Type
File Version
Parameter Description
currently, a single file type is recognized, named Wide-Field Time-Gated Data
for a file compliant with this version, the string would be 0.1
DAQ Parameters cluster (saved as a Table)
Metadata string
Gate Images 3-dimensional array of single precision floating point numbers
A snapshot of a file as viewed with HDFView, the free HDF5 file viewer developed by the HDF Group is shown below: