Skip to content

mindsdb/mindsdb_python_sdk

Folders and files

NameName
Last commit message
Last commit date
Oct 24, 2024
Oct 24, 2024
Apr 26, 2024
Sep 6, 2024
Nov 15, 2024
Oct 25, 2024
Apr 5, 2021
Oct 16, 2020
Jun 9, 2023
Nov 11, 2024
Jun 6, 2023
Sep 22, 2023
Jul 23, 2024
Apr 12, 2024
Jun 6, 2023

Repository files navigation

Python MindsDB SDK

The Python MindsDB SDK allows you to connect to a MindsDB server from Python using the HTTP API.

Installation

pip install mindsdb_sdk

Example

Connecting to the MindsDB server

You can establish a connection to the MindsDB server using the SDK. Here are some examples:

Connect to a local MindsDB server

import mindsdb_sdk
con = mindsdb_sdk.connect()
con = mindsdb_sdk.connect('http://127.0.0.1:47334')

Connect to the MindsDB Cloud

import mindsdb_sdk
con = mindsdb_sdk.connect(login='[email protected]', password='-')
con = mindsdb_sdk.connect('https://cloud.mindsdb.com', login='[email protected]', password='-')

Connect to a MindsDB Pro server

import mindsdb_sdk
con = mindsdb_sdk.connect('http://<YOUR_INSTANCE_IP>', login='[email protected]', password='-', is_managed=True)

Basic usage

Once connected to the server, you can perform various operations. Here are some examples:

# Get a list of databases
databases = con.databases.list()

# Get a specific database
database = databases[0]  # Database type object

# Perform an SQL query
query = database.query('select * from table1')
print(query.fetch())

# Create a table
table = database.tables.create('table2', query)

# Get a project
project = con.projects.proj

# or use mindsdb project
project = con

# Perform an SQL query within a project
query = project.query('select * from database.table join model1')

# Create a view
view = project.views.create('view1', query=query)

# Get a list of views
views = project.views.list()
view = views[0]
df = view.fetch()

# Get a list of models
models = project.models.list()
model = models[0]

# Use a model for prediction
result_df = model.predict(df)
result_df = model.predict(query)

# Create a model
timeseries_options = {
    'order': 'date',
    'window': 5,
    'horizon': 1
}
model = project.models.create(
    'rentals_model',
    predict='price',
    query=query,
    timeseries_options=timeseries_options
)

# Describe a model
model.describe()

You can find more examples in this Google colab notebook

Examples

https://github.com/mindsdb/mindsdb_python_sdk/tree/staging/examples

API Documentation

The API documentation for the MindsDB SDK can be found at https://mindsdb.github.io/mindsdb_python_sdk/.

Generating API docs locally:

cd docs
pip install -r requirements.txt
make html

The online documentation is automatically updated by pushing changes to the docs branch.

Testing

To run all the tests for the components, use the following command:

env PYTHONPATH=./ pytest

Contributing

We welcome contributions to the MindsDB SDK. If you'd like to contribute, please refer to the contribution guidelines for more information.

License

The MindsDB SDK is licensed under the MIT License. Feel free to use and modify it according to your needs