GraphTable API Reference¶

Type: groggy.GraphTable

Overview¶

Tabular representation of graph data with separate nodes and edges tables.

Primary Use Cases: - Exporting graph data to CSV/Parquet - Converting to pandas DataFrames - Tabular analysis of graph data

Related Objects: - Graph - Subgraph - NodesTable - EdgesTable

Complete Method Reference¶

The following methods are available on GraphTable objects. This reference is generated from comprehensive API testing and shows all empirically validated methods.

Method	Returns	Status
`auto_assign_edge_ids()`	`GraphTable`	✓
`edges()`	`EdgesTable`	✓
`from_federated_bundles()`	`?`	✗
`get_bundle_info()`	`?`	✗
`head()`	`GraphTable`	✓
`is_empty()`	`bool`	✓
`load_bundle()`	`?`	✗
`merge()`	`?`	✗
`ncols()`	`int`	✓
`nodes()`	`NodesTable`	✓
`nrows()`	`int`	✓
`save_bundle()`	`?`	✗
`shape()`	`tuple`	✓
`stats()`	`dict`	✓
`tail()`	`GraphTable`	✓
`to_edges()`	`?`	✗
`to_graph()`	`Graph`	✓
`to_nodes()`	`?`	✗
`to_subgraphs()`	`?`	✗
`validate()`	`str`	✓
`verify_bundle()`	`?`	✗
`viz()`	`VizAccessor`	✓

Legend: - ✓ = Method tested and working - ✗ = Method failed in testing or not yet validated - ? = Return type not yet determined

Detailed Method Reference¶

Creating GraphTable¶

GraphTables are created from graphs or subgraphs:

import groggy as gr

g = gr.generators.karate_club()

# From graph
table = g.table()

# From subgraph
young = g.nodes[g.nodes["age"] < 30]
young_table = young.table()

# Load from bundle
table = gr.GraphTable.load_bundle("graph.bundle")

Accessing Components¶

`nodes`¶

Access the nodes table.

Returns: - NodesTable: Table of node data

Example:

table = g.table()
nodes_table = table.nodes

# View nodes
print(nodes_table.head())

`edges`¶

Access the edges table.

Returns: - EdgesTable: Table of edge data

Example:

table = g.table()
edges_table = table.edges

# View edges
print(edges_table.head())

Properties¶

`shape()`¶

Get dimensions (total rows, total columns).

Returns: - tuple[int, int]: (rows, cols) across nodes and edges

Example:

table = g.table()
rows, cols = table.shape()
print(f"Total: {rows} rows, {cols} columns")

`nrows()`¶

Get total number of rows.

Returns: - int: Number of rows

Example:

table = g.table()
print(f"{table.nrows()} total rows")

`ncols()`¶

Get total number of columns.

Returns: - int: Number of columns

Example:

table = g.table()
print(f"{table.ncols()} total columns")

`is_empty()`¶

Check if table has no data.

Returns: - bool: True if empty

Example:

table = g.table()
if table.is_empty():
    print("No data")

Display Methods¶

`head(n=5)`¶

Show first n rows.

Parameters: - n (int): Number of rows (default 5)

Returns: - GraphTable: Table with first n rows

Example:

table = g.table()
table.head()      # First 5 rows
table.head(10)    # First 10 rows

`tail(n=5)`¶

Show last n rows.

Parameters: - n (int): Number of rows (default 5)

Returns: - GraphTable: Table with last n rows

Example:

table = g.table()
table.tail()      # Last 5 rows
table.tail(10)    # Last 10 rows

`stats()`¶

Get summary statistics.

Returns: - dict: Statistics about the table

Example:

table = g.table()
stats = table.stats()
print(stats)
# {'nodes': 34, 'edges': 78, 'node_attrs': ['age', 'club'], ...}

Validation¶

`validate()`¶

Validate table structure.

Returns: - str: Validation report

Example:

table = g.table()
report = table.validate()
print(report)
# "Valid: 34 nodes, 78 edges, all edge endpoints exist"

Notes: - Checks that edge endpoints exist in nodes table - Validates data types - Reports any inconsistencies

Conversion¶

`to_graph()`¶

Convert table back to Graph.

Returns: - Graph: New graph from table data

Example:

table = g.table()

# Modify table data (via pandas, etc.)
# ...

# Convert back to graph
g2 = table.to_graph()

Use cases: - Round-trip processing (Graph → Table → modify → Graph) - Building graphs from table data - Restoring from saved bundles

I/O Operations¶

`save_bundle(path)`¶

Save complete graph to bundle file.

Parameters: - path (str): File path to save to

Returns: - None

Example:

table = g.table()
table.save_bundle("my_graph.bundle")

Notes: - Bundles store complete graph state - Efficient binary format - Preserves all attributes and structure

`load_bundle(path)` (class method)¶

Load graph from bundle file.

Parameters: - path (str): File path to load from

Returns: - GraphTable: Loaded table

Example:

table = gr.GraphTable.load_bundle("my_graph.bundle")
g = table.to_graph()

Utility Methods¶

`auto_assign_edge_ids()`¶

Automatically assign IDs to edges if missing.

Returns: - GraphTable: Table with edge IDs assigned

Example:

# If edges lack IDs
table_with_ids = table.auto_assign_edge_ids()

`viz()`¶

Access visualization methods.

Returns: - VizAccessor: Visualization accessor

Example:

viz = table.viz()
# Use viz methods

Usage Patterns¶

Pattern 1: Export to Pandas¶

table = g.table()

# Get nodes as DataFrame
nodes_df = table.nodes.to_pandas()
print(nodes_df.head())

# Get edges as DataFrame
edges_df = table.edges.to_pandas()
print(edges_df.head())

# Combined analysis
import pandas as pd
print(f"Nodes: {len(nodes_df)}")
print(f"Edges: {len(edges_df)}")
print(f"Attributes: {list(nodes_df.columns)}")

Pattern 2: Save/Load Graph¶

# Save
g = gr.generators.karate_club()
table = g.table()
table.save_bundle("karate.bundle")

# Load later
table2 = gr.GraphTable.load_bundle("karate.bundle")
g2 = table2.to_graph()

# Verify
assert g2.node_count() == g.node_count()
assert g2.edge_count() == g.edge_count()

Pattern 3: Round-Trip Processing¶

# Graph to table
table = g.table()

# Export to pandas for processing
df = table.nodes.to_pandas()

# Modify
df['age_group'] = df['age'].apply(lambda x: 'young' if x < 30 else 'old')

# Save modified data
df.to_csv("nodes_modified.csv", index=False)

# Rebuild graph
import pandas as pd
nodes_df = pd.read_csv("nodes_modified.csv")
# Build new graph from modified data

Pattern 4: Validation¶

table = g.table()

# Check validity
report = table.validate()
print(report)

# Check if empty
if not table.is_empty():
    stats = table.stats()
    print(f"Valid graph: {stats}")

Pattern 5: Inspection¶

table = g.table()

# Quick overview
print(f"Shape: {table.shape()}")
print(f"Rows: {table.nrows()}, Cols: {table.ncols()}")

# Preview data
print("\nFirst 5 rows:")
table.head()

print("\nLast 5 rows:")
table.tail()

# Detailed stats
stats = table.stats()
for key, value in stats.items():
    print(f"{key}: {value}")

Quick Reference¶

Properties¶

Method	Returns	Description
`shape()`	`tuple`	(rows, cols)
`nrows()`	`int`	Total rows
`ncols()`	`int`	Total columns
`is_empty()`	`bool`	Has data?

Components¶

Method	Returns	Description
`nodes`	`NodesTable`	Nodes table
`edges`	`EdgesTable`	Edges table

Display¶

Method	Description
`head(n)`	First n rows
`tail(n)`	Last n rows
`stats()`	Summary statistics

I/O¶

Method	Description
`save_bundle(path)`	Save to file
`load_bundle(path)`	Load from file
`to_graph()`	Convert to Graph

Relationship with Other Tables¶

GraphTable is a container for NodesTable and EdgesTable:

table = g.table()           # GraphTable

nodes_table = table.nodes   # NodesTable
edges_table = table.edges   # EdgesTable

# Each has export methods
nodes_df = nodes_table.to_pandas()
edges_df = edges_table.to_pandas()

nodes_table.to_csv("nodes.csv")
edges_table.to_parquet("edges.parquet")

Key difference: - GraphTable: Complete graph representation (nodes + edges) - NodesTable: Only node data - EdgesTable: Only edge data

Object Transformations¶

GraphTable can transform into:

GraphTable → NodesTable: table.nodes
GraphTable → EdgesTable: table.edges
GraphTable → DataFrame: table.to_pandas()
GraphTable → Files: table.to_csv(), table.to_parquet()

See Object Transformation Graph for complete delegation chains.

Additional Methods¶

`edges()`¶

Edges.

Returns: - EdgesTable: Return value

Example:

obj.edges()

`from_federated_bundles(bundle_paths)`¶

From Federated Bundles.

Parameters: - bundle_paths: bundle paths

Returns: - None: Return value

Example:

obj.from_federated_bundles(bundle_paths=...)

`get_bundle_info(bundle_path)`¶

Get Bundle Info.

Parameters: - bundle_path: bundle path

Returns: - None: Return value

Example:

obj.get_bundle_info(bundle_path=...)

`merge(tables)`¶

Merge.

Parameters: - tables: tables

Returns: - None: Return value

Example:

obj.merge(tables=...)

`nodes()`¶

Nodes.

Returns: - NodesTable: Return value

Example:

obj.nodes()

`to_edges()`¶

To Edges.

Returns: - None: Return value

Example:

obj.to_edges()

`to_nodes()`¶

To Nodes.

Returns: - None: Return value

Example:

obj.to_nodes()

`to_subgraphs()`¶

To Subgraphs.

Returns: - None: Return value

Example:

obj.to_subgraphs()

`verify_bundle(bundle_path)`¶

Verify Bundle.

Parameters: - bundle_path: bundle path

Returns: - None: Return value

Example:

obj.verify_bundle(bundle_path=...)

GraphTable API Reference¶

Overview¶

Complete Method Reference¶

Detailed Method Reference¶

Creating GraphTable¶

Accessing Components¶

nodes¶

edges¶

Properties¶

shape()¶

nrows()¶

ncols()¶

is_empty()¶

Display Methods¶

head(n=5)¶

tail(n=5)¶

stats()¶

Validation¶

validate()¶

Conversion¶

to_graph()¶

I/O Operations¶

save_bundle(path)¶

load_bundle(path) (class method)¶

Utility Methods¶

auto_assign_edge_ids()¶

viz()¶

Usage Patterns¶

Pattern 1: Export to Pandas¶

Pattern 2: Save/Load Graph¶

Pattern 3: Round-Trip Processing¶

Pattern 4: Validation¶

Pattern 5: Inspection¶

Quick Reference¶

Properties¶

Components¶

Display¶

I/O¶

Relationship with Other Tables¶

Object Transformations¶

See Also¶

Additional Methods¶

edges()¶

from_federated_bundles(bundle_paths)¶

get_bundle_info(bundle_path)¶

merge(tables)¶

nodes()¶

to_edges()¶

to_nodes()¶

to_subgraphs()¶

verify_bundle(bundle_path)¶

`nodes`¶

`edges`¶

`shape()`¶

`nrows()`¶

`ncols()`¶

`is_empty()`¶

`head(n=5)`¶

`tail(n=5)`¶

`stats()`¶

`validate()`¶

`to_graph()`¶

`save_bundle(path)`¶

`load_bundle(path)` (class method)¶

`auto_assign_edge_ids()`¶

`viz()`¶

`edges()`¶

`from_federated_bundles(bundle_paths)`¶

`get_bundle_info(bundle_path)`¶

`merge(tables)`¶

`nodes()`¶

`to_edges()`¶

`to_nodes()`¶

`to_subgraphs()`¶

`verify_bundle(bundle_path)`¶