Cryptostore

* Feature: Support for Coingecko aggregated data per coin, to be used with a new data channel 'market_info'
* Feature: Config.yaml supports individual exchange's Feed class params.
* Feature: Support for 'http_proxy' in exchanges config.
* Bugfix: Update aggregator util functions for level 2 and level 3 order book to handle book delta
* Update: Remove deprecated code.
* Bugfix: Update config parsing to account for changes to cryptofeed

* Bugfix: Fix book building example
* Feature: Ability to pull config file from S3 bucket instead of local filesystem, polling for updated config is supported.
* Feature: Updated to work with latest version of cryptofeed
* Feature: Liquidations support for Arctic
* Bugfix: Aggregating parquet files at the end of time no longer causes file not found error.
* Feature: Temporary parquet files are now gathering in a seperate folder
* Bugfix: Parquet column type fix
* Bugfix: Fix issue where max depth was being passed incorrectly to cryptofeed
* Bugfix: rename kafka host kwarg to bootstrap, as cryptofeed backend expects
* Feature: Candles support
* Bugfix: Fix l2/l3_book_flatten routines
* Feature: Update to work with latest cryptofeed version/changes. Update InfuxDB backend to work with 2.x. Drop support for InfluxDB 1.x

* Feature: Influxdb 1.x authentication support
* Bugfix: Extend timeout of Google Drive connection
* Feature: Parquet file compression
* Feature: Parquet file content optimization
* Feature: Option to enable parquet file appending
* Feature: Retry writes to storage engines (rather than dying).
* Feature: Support for snapshot_interval
* Feature: Parquet files can be stored in directories per day
* Bugfix: Set arctic storage quota to unlimited (rather than 10G)
* Bugfix: Regression in arctic storage backend
* Feature: Opened parquet files (when appended) are suffixed with '.tmp'
* Feature: Liquidations support

* Feature: Config options for controlling data channel timeouts
* Feature: Path option in config for local parquet storage
* Feature: Configurable filenames for parquet storage
* Bugfix: Ticker missing from ZMQ pass through
* Feature: Store data to Google Drive
* Bugfix: Kafka now supports the time interval for storage intervals (feature added in 0.2.1)

* Feature: Open Interest support
* Example: Added example code for orderbook reconstruction from data in Arctic
* Feature: More granular control over data storage intervals
* Feature: Support receipt timestamps from cryptostore
* Example: Added example code for ZMQ passthrough

* Bugfix: Fixed elasticsearch timestamps
* Feature: Support for different S3 endpoints to allow writing to other providers
* Bugfix: Missing config value for redis socket