Data exploration with Python and Visplore

The Python integration is a key feature of every Visplore version, and can be used for free!

The Python integration enables data scientists to work with Python and Visplore simultaneously, combining the strengths of both tools:

Doing interactive visual data exploration and interactive data preparation in Visplore while using Python for unlimited analytics features and for creating operational data processing pipelines. This is perfectly in line with CRISP-DM, a widely adopted process model of data science projects:

Using Visplore for gaining a rapid understanding of the data and (to some extent) preparing it, and using Python for modeling and operational deployment.

Get started with Visplore Free

Key functionality offered by this API includes:

Opening a new Visplore instance and establishing a communication channel between the Python environment and Visplore
Sending Python data structures such as Pandas data frames to Visplore, possibly extending or replacing previously sent data
Getting user-defined data selections or labels from Visplore to Python to be stored in numpy vectors.
Retrieving data subsets from Visplore as Pandas data frames
Retrieving results such as Pivot tables or graphics from Visplore in a way that is easy to automatize

The integration is based on a Python API that can be used in any Python environment, for example Jupyter Notebook or Spyder, but also plain Python.

Using this API opens up multiple exciting usage scenarios:

Exploration

Having a Pandas data frame, just send it to Visplore using a single line of code (“send_data(dataframe)”). Within seconds, the pre-configured cockpits will tell you everything about distributions, missing values, trends, correlations, outliers, and so on – highly interactive and without any further need for scripting the graphics by hand

Feature engineering

Build new features using Python code and tune their parameters in Python while keeping the results synchronized with Visplore. This integrated workflow speeds up the validation dramatically and allows you to engineer better features faster.

Data subset selection and labeling

Use the intuitive data selection of Visplore to label plausible data as training data, validation data, or anomalies and much more. Then, use these labels to train and validate models in Python based on meaningful data.

Feature selection

Select features in Visplore which are actually meaningful and contain plausible data, also avoiding sets of highly correlated features. Then query this feature set from Python for working with a plausible and informative set.

Interactive data cleaning

Data can be cleaned interactively in Visplore. For example, users may easily remove outliers or fill gaps in time series with interpolated or predicted values. The cleaned data table can be retrieved from Python as Pandas data frame.

Reporting

Use Visplore for generating Visplore graphics in a fully automated way, for example for reporting. All necessary steps from opening Visplore to exporting until closing can be specified as a Python script without any manual intervention.

Connecting additional data source

Use Python as a “bridge” for connecting data sources to Visplore for which no built-in connectivity exists so far.

In a nutshell, using Python and Visplore together enables data scientists to preserve all flexibility and power of Python while opening up fast, interactive and intuitive visual data exploration and preparation especially for time series data.

Thomas Mühlbacher, Visplore CTO

Get Visplore to take your data visualization
with Python to the next level.

Get Free Version Book Live Demo

Data exploration with Python and Visplore

Key functionality offered by this API includes:

Using this API opens up multiple exciting usage scenarios:

Exploration

Feature engineering

Data subset selection and labeling

Feature selection

Interactive data cleaning

Reporting

Connecting additional data source

Get Visplore to take your data visualization
with Python to the next level.

Why Visplore?

Resources

Newsletter

Data exploration with Python and Visplore

Key functionality offered by this API includes:

Using this API opens up multiple exciting usage scenarios:

Exploration

Feature engineering

Data subset selection and labeling

Feature selection

Interactive data cleaning

Reporting

Connecting additional data source

Get Visplore to take your data visualizationwith Python to the next level.

Why Visplore?

Resources

Newsletter

Get Visplore to take your data visualization
with Python to the next level.