First draft of numfocus_onepager.

numfocus_onepager.md

@@ -0,0 +1,80 @@
+# NetworkX
+
+## Summary
+
+NetworkX is a Python package for the creation, manipulation, and study of the
+structure, dynamics, and functions of complex networks.
+NetworkX provides intuitive data structures for describing networks (or graphs)
+with the flexibility to support a wide range of relational data.
+The library also provides an extensive set of tools including standard graph
+algorithms; network structure and analysis measures; and utilities for
+creating a wide variety of graphs, either from existing data or with
+well-known generation methods.
+The library is written in pure Python, with an emphasis on readable,
+well-documented, and thoroughly-tested code.
+Graph and network analysis are ubiquitous with applications spanning many
+domains: from mathematics and theoretical computer science to astronomy,
+linguistics, biology, social science, and beyond.
+NetworkX is the fundamental package for network analysis in the scientific
+Python ecosystem, supporting exploration and applications across all domains.
+
+## Use Cases / Applications
+
+NOTE: these are pretty generic - having some more specific examples would be
+good.
+
+ - Analysis of social networks such as Twitter, Facebook, or YouTube; including
+   community detection, information flow, and suggestion and recommendation
+   algorithms.
+ - Biology and public health: from protein folding to contact tracing.
+ - Computer science and information theory - many fundamental data structures
+   and algorithms are described in terms of graphs, from tree structures and
+   information codings to search and path finding algorithms.
+
+Other ideas from perusal of dependents on GitHub:
+ - AI/ML/neural networks
+   * [netket](https://github.com/netket/netket)
+   * [Deep graph library (DGL)](https://github.com/dmlc/dgl)
+   * [Maze RL framework](https://github.com/enlite-ai/maze)
+   * [allen act (embodied AI)](https://github.com/allenai/allenact)
+   * [evalml (automl framework)](https://github.com/alteryx/evalml)
+   * [CompilerGym (RL for compiler optimzations from FB)](https://github.com/facebookresearch/CompilerGym)
+   * [open_spiel - RL for games](https://github.com/deepmind/open_spiel)
+   * [cogDL (DL with graphs)](https://github.com/THUDM/cogdl)
+   * [AWS autogluon](https://github.com/awslabs/autogluon)
+ - neuroscience
+   * [neurokernel (fruitfly brain models)](https://github.com/neurokernel/neurokernel)
+ - Geospatial analysis
+   * [spopt](https://github.com/pysal/spopt)
+ - microbiology
+   * [biotite](https://github.com/biotite-dev/biotite)
+ - Quantum computing
+   * [Qiskit](https://github.com/Qiskit/qiskit-ignis)
+ - Genomics
+   * [RagTag](https://github.com/malonge/RagTag)
+   * [GenomeWorks](https://github.com/clara-parabricks/GenomeWorks)
+ - Data visualization
+   * [dtale (flask project for pandas viz)](https://github.com/man-group/dtale)
+ - NLP
+   * [hanlp](https://github.com/hankcs/HanLP)
+   * [haystack](https://github.com/deepset-ai/haystack)
+
+## Planned Features
+
+ - Improved interoperability with other scientific Python libraries. This
+   includes tighter integration with foundational packages like NumPy, SciPy,
+   and Matplotlib for improved performance, as well as domain-specific
+   libraries in which network analysis plays a central role (e.g. Geopandas).
+
+ - Increased selection of algorithms, including Louvain community detection,
+   VF2++ subgraph isomorphism, and heuristic-based approaches for solving the
+   traveling salesman problem.
+
+ - More flexible graph generators to improve scalability and better support
+   non-NetworkX-specific data structures.
+
+## Project needs
+
+ - Graph visualization performance and scalability | 2000 hours
+ - Exploring interface with other computation backends | 2000 hours
+ - Community manager (part time) | 250 hours / year