The dataset viewer is not available for this subset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 82, in _split_generators
raise ValueError(
ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 65, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Dataset Details
Dataset Description
This dataset contains causal sets of types
- layered
- embedded in polynomial monifold with complex cuts (x2)
- embedded in polynomial manifold (x2)
- random (random DAG fullfilling the cset requirements)
- grid-like with different cells (hexagonal, rhombic, triangular, rectangular, rhombic, oblique)
- merged (random and polynomial)
- destroyed (polynomial dataset with some edges being flipped randomly)
- merged ambiguous (too small to decide between random and polynomial manifold)
- destroyed ambiguous (too little difference to decide between random and polynomial manifold)
Data has been created with QuantumGrav.jl based on Causalsets.jl
Uses
Used to train models for classifying causal sets, or for representation learning on causal sets. Other use cases have not been explored.
Dataset Structure
2048_5k.tar.gzip2:
- for each type: 5000 causal sets of cardinality 512 to 2048 uniformly sampled
- a zarr v2 store for each type of causal sets
- a .yaml config file for each that defines all parameters used to create the data. This includes data about the git commit and branch used.
- a julia script
create_data.jlthat has been used to create the data - a julia script
augment_data.jlthat has been used to create the augmented data
Each causal set is stored as a group in the zarr store. Each group contains the following arrays and attributes:
adjacency_matrix: Adjacency matrix of the causal setlink_matrix: Link matrix (transitive reduction of adjacency matrix)- in- and out degree for adj. and link matrix
- chain number histogram
- causal set type (as integer)
- manifold-likeness (as bool)
- number of zero degrees (in and out, for adj. and link matrix)
- relation numbers
- relation dimension
- cardinality histogram (spectrum)
- maximum forward- and backward path length in the graph represented by the link matrix for each node. These are called
max_pathlen_forwardandmax_pathlen_backward.
All properties of the dataset have been uniformly sampled. See the config.yaml files for details.
Additionally, parallel coordinates plots for the sampled properties are provided.
2048_5k_normaldist.tar.gzip2:
This is largely the same as 2048_5k.zip, but the properties of the causal sets have been drawn from normal distributions instead of uniform distributions. Note:
- There is no deterministic way to assure (without additional modifications of the sampling code) that the ambiguous types are always within the same range as the non-ambiguous types when drawing from normal distributions. Hence, what is labeled as ambiguous in this dataset may not be ambiguous in the uniform distribution dataset, and vice versa, and there is some overlap between ambiguous and non-ambiguous types.
- the path length node features have been renamed:
max_pathlen_forward->max_pathlen_forward_link,max_pathlen_backward->max_pathlen_backward_linkfor clarity.
Personal and Sensitive Information
None
Bias, Risks, and Limitations
This data contains only causal sets which's properties are drawn from uniform distributions (or normal, where applicable). Causal set structures can change when changing the distributions of their underlying properties, especially for small causal sets. This might limit the applicability of models trained on this data to datasets drawn from vastly different distributions or when using vastly largere causal sets.
Dataset Card Contact
- Downloads last month
- 16