Concrete ExploreDataset backend over a parquet directory

The exploreDElight backend. Reads the parquet directory a native Rust process writes from a canonical exploreDE/1.0 .h5ad, and satisfies the same accessor contract as every other backend. Inherits all config derivation and DE assembly from ExploreDatasetH5AD; only the loader differs.

Super classes

ExploreDataset -> ExploreDatasetH5AD -> ExploreDatasetParquet

Methods

Public methods

• ExploreDatasetParquet$new()

• ExploreDatasetParquet$backend_kind()

• ExploreDatasetParquet$manifest()

• ExploreDatasetParquet$source_dir()

• ExploreDatasetParquet$verify_source()

• ExploreDatasetParquet$all_pathways()

• ExploreDatasetParquet$clone()

Inherited methods

• ExploreDataset$assay_label()
• ExploreDataset$background_expression()
• ExploreDataset$contrast_label()
• ExploreDataset$contrast_meta()
• ExploreDataset$contrasts()
• ExploreDataset$data_type()
• ExploreDataset$de_all()
• ExploreDataset$de_display_extras()
• ExploreDataset$enrichment_filter()
• ExploreDataset$enrichment_kinds()
• ExploreDataset$factor_levels()
• ExploreDataset$factor_names()
• ExploreDataset$feature_display()
• ExploreDataset$feature_id_kind()
• ExploreDataset$feature_ids()
• ExploreDataset$format()
• ExploreDataset$has_assay()
• ExploreDataset$has_de()
• ExploreDataset$has_enrichment()
• ExploreDataset$has_pvalue()
• ExploreDataset$n_features()
• ExploreDataset$n_samples()
• ExploreDataset$omics_class()
• ExploreDataset$organism()
• ExploreDataset$primary_assay()
• ExploreDataset$primary_assay_name()
• ExploreDataset$primary_contrast()
• ExploreDataset$primary_grouping()
• ExploreDataset$proteomics_choices()
• ExploreDataset$raw_assay()
• ExploreDataset$raw_assay_name()
• ExploreDataset$ref_build()
• ExploreDataset$sample_factor()
• ExploreDataset$sample_ids()
• ExploreDataset$secondary_grouping()
• ExploreDataset$sgrna_to_gene()
• ExploreDataset$source()
• ExploreDatasetH5AD$assay()
• ExploreDatasetH5AD$assay_names()
• ExploreDatasetH5AD$backend()
• ExploreDatasetH5AD$de()
• ExploreDatasetH5AD$enrichment()
• ExploreDatasetH5AD$enrichment_input()
• ExploreDatasetH5AD$enrichment_table()
• ExploreDatasetH5AD$features()
• ExploreDatasetH5AD$samples()

---

ExploreDatasetParquet$new()

Read the parquet directory (manifest, obs, var, layers, varm, uns), build the config, prepare the feature and DE tables, and warn when the source .h5ad has changed since export.

Usage

ExploreDatasetParquet$new(input)

Arguments

input

Path to the parquet directory.

---

ExploreDatasetParquet$backend_kind()

Returns "parquet".

Usage

ExploreDatasetParquet$backend_kind()

---

ExploreDatasetParquet$manifest()

The parsed manifest.json of the parquet directory.

Usage

ExploreDatasetParquet$manifest()

---

ExploreDatasetParquet$source_dir()

Path of the parquet directory.

Usage

ExploreDatasetParquet$source_dir()

---

ExploreDatasetParquet$verify_source()

Compare the md5 of the source .h5ad with the one recorded in the manifest: TRUE when it matches, FALSE when not, NA when there is nothing to check against.

Usage

ExploreDatasetParquet$verify_source(path = NULL)

Arguments

path

Path to the source .h5ad; NULL (default) uses the path recorded in the manifest.

---

ExploreDatasetParquet$all_pathways()

GO terms as in the base class, but an empty vector when clusterProfiler is not installed.

Usage

ExploreDatasetParquet$all_pathways()

---

ExploreDatasetParquet$clone()

The objects of this class are cloneable with this method.

Usage

ExploreDatasetParquet$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.