Docker compose for data analysis
- ClickHouse is an open source column-oriented database management system capable of real time generation of analytical data reports using SQL queries.
- Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text.
- TABIX SQL Editor & Open source simple business intelligence for Clickhouse.
- ClickHouse use 8123 port for http clients and 9000 for console client
- Jupyter use 8888
- TABIX use 8080
Jupyter Uses as image https://hub.docker.com/r/jupyter/datascience-notebook/
To use password auth set command at docker-compose.yml with pwd hash generated by IPython.lib.passwd()
Set volume /home/jovyan/work to you data dir, or use default
Clickhouse Build from Dockerfile on Ubuntu base image.
Change clickhouse/config.xml and clickhouse/users.xml in accordance with the documentation https://clickhouse.yandex/
Set user password in clickhouse/users.xml
Set volume to datastore in docker-compose.yml
Tabix Uses as image
To use basic HTTP auth set environment variable in docker-compose.yml
In work directory of Jupyter Notebook added clickhouse python connection helper. Set USER/PASSWORD or use default readonly user. Import to you *.ipynb file clickhouse_helper and send query. You can get raw responce or pandas dataframe.