Biocypher

What's Changed
* Optional-id by slobentanzer in https://github.com/biocypher/biocypher/pull/212
* draft pandas-tutorial in jupyter by dbdimitrov in https://github.com/biocypher/biocypher/pull/209


**Full Changelog**: https://github.com/biocypher/biocypher/compare/v0.5.14...v0.5.16

Fixed configuration mistake that prevented log level of the file handler to be "debug".

Functions:
- `bc.add()` to add nodes and edges to the in-memory store
- `bc.to_df()` to return a dictionary of individual Pandas dataframes, one per type of node/edge

Also added deduplicator module and class to share this functionality between writer and Pandas modules.

This refactor introduces batch writing capabilities for PostgreSQL and ArangoDB. This support is experimental and will be refined in coming releases.

This release introduces a breaking change in the configuration part of BioCypher. The organisation of the YAML configuration file has changed, compartmentalising settings into `biocypher` core settings and other, DBMS-specific settings. Please refer to https://biocypher.org/installation.html#configuration for more info.

The refactoring modularises the architecture of BioCypher. The interface function that was previously entangled with Neo4j Driver functionality has been moved to the new core class `BioCypher`. This class organises the interplay between input sources, ontologies, and output formats, guided by the user configuration. It can request offline file writers as well as online DBMS connectors, which it then uses to write the input data stream based on the ontology mapping defined by the user.

The previous `Driver` module is still usable, but simply redirects to the new `BioCypher` interface class and gives a deprecation warning. Some parameters are not available any more and need to be configured in the `biocypher_config.yaml`.

In addition, small utility and usability changes have been introduced.