DBRepo Python Library

Official client library for DBRepo, a database repository to support research based on requests, pydantic and pika.

Installing

$ python -m pip install dbrepo

This package supports Python 3.11+.

Quickstart

Get public data from a table as pandas DataFrame:

from dbrepo.RestClient import RestClient

client = RestClient(endpoint="https://test.dbrepo.tuwien.ac.at")
# Get a small data slice of just three rows
df = client.get_table_data(database_id="e0d82287-9099-4077-8f69-3c19fc3bc145",
                           table_id="71f8c746-ea26-4651-b3f2-ce46830f1af4",
                           page=0, size=3)
print(df)
#     x_coord         component   unit  ... value stationid meantype
# 0  16.52617  Feinstaub (PM10)  µg/m³  ...  21.0   01:0001      HMW
# 1  16.52617  Feinstaub (PM10)  µg/m³  ...  23.0   01:0001      HMW
# 2  16.52617  Feinstaub (PM10)  µg/m³  ...  26.0   01:0001      HMW
#
# [3 rows x 12 columns]

Create table and import DataFrame into a table:

import pandas as pd
from dbrepo.RestClient import RestClient

client = RestClient(endpoint="https://test.dbrepo.tuwien.ac.at", username="foo",
                    password="bar")
df = pd.DataFrame(data={'x_coord': 16.52617, 'component': 'Feinstaub (PM10)',
                        'unit': 'µg/m³', ...})
df = df.set_index(['x_coord'])
client.create_table(database_id="e0d82287-9099-4077-8f69-3c19fc3bc145",
                    name="Sensor", is_public=True, is_schema_public=True,
                    dataframe=df)

... or just create the table schema by setting create_table(..., withData=False).

In both cases it is important to set the index to existing columns that uniquely identify a row. You can specify multiple columns:

...
df = df.set_index(['some_column', 'some_other_column'])

Supported Features & Best-Practices

• Manage user account (docs)
• Manage databases (docs)
• Manage database access & visibility (docs)
• Import dataset (docs)
• Create persistent identifiers (docs)
• Execute queries (docs)
• Get data from tables/views/subsets

Configure

All credentials can optionally be set/overridden with environment variables. This is especially useful when sharing Jupyter Notebooks by creating an invisible .env file and loading it:

REST_API_ENDPOINT="https://dbrepo1.ec.tuwien.ac.at"
REST_API_USERNAME="foo"
REST_API_PASSWORD="bar"
REST_API_SECURE="True"
AMQP_API_HOST="https://dbrepo1.ec.tuwien.ac.at"
AMQP_API_PORT="5672"
AMQP_API_USERNAME="foo"
AMQP_API_PASSWORD="bar"
AMQP_API_VIRTUAL_HOST="dbrepo"

You can disable logging by setting the log level to e.g. INFO:

from dbrepo.RestClient import RestClient
import logging
logging.getLogger().setLevel(logging.INFO)
...
client = RestClient(...)

Roadmap

• Searching

Contact

• Prof. Andreas Rauber^{TU Wien}
• DI. Martin Weise^{TU Wien}