Skip to content

Roadmap

Jordan Matelsky edited this page Oct 14, 2020 · 2 revisions

Features that will exist, someday.

Status Description
🤔 Currently under development
No plans to implement yet

🤔 Derived-attribute filters

Summary

Attributes derived from the graph, not from the data (e.g. degree, betweenness centrality, etc)

Proposed Notation(s)

Prefix derived attributes with !:

A.!degree > 5

Bracket notation:

A -> B 

degree{my_node} > 5

🤔 N-hop relations

Summary

To represent an unknown number of intermediate nodes before reaching a final destination.

Proposed Notation(s)

All examples represent the following, where dont_care_B and dont_care_C are omitted:

A -> dont_care_B
dont_care_B -> dont_care_C
dont_care_C -> D

Parentheses

A -(2)-> D

With attributes:

A -(2 [weight > 4])-> D

⏳ Include edge/node attributes in result

Summary

When making a query, allow the user to notate that they want the query to also return the edge/node metadata attributes that satisfy the constraints:

X -> Y [weight > 4]
X.type != excitatory

Here, we want to notate that we care about X.type and [XY].weight, and that the returned table should look like:

X Y [XY].weight X.type

Instead of just containing X and Y headers.

Proposed Notation(s)

Standalone declaration lines

Desired attributes can be signified with an asterisk:

X -> Y [weight > 4]
X.type != excitatory

+[XY].weight
+X.type