Skip to main content
Version: Next

Running the Radar Pipeline in Jupyter Notebooks or as a python library

To provide a more interactive experience for users, you can run the RADAR-pipeline in Jupyter notebooks. This allows you to visualize the data processing steps and results in real-time.

To get started, you need to install the necessary dependencies and set up your Jupyter notebook environment.

Step 1. Install Jupyter Notebook and RADAR-pipeline​

You can install Jupyter Notebook and the RADAR-pipeline module using pip. Open your terminal and run the following commands:

pip install jupyter
pip install radarpipeline

Step 2. Create a Jupyter Notebook​

Create a new Jupyter notebook file (e.g., radar_pipeline_notebook.ipynb) and open it in Jupyter Notebook.

Step 3. Import RADAR-pipeline in the Notebook​

In your Jupyter notebook, you can import the RADAR-pipeline module and start using it. Here is an example of how to set up a basic pipeline:

from radarpipeline import radarpipeline

radarpipeline module have several handy functions to help you set up, read the data and run any pipeline.

To read and process data, you can use the following code snippet:

variables = ['android_app_battery_level', 'android_app_step_count']
datas = radarpipeline.read(PATH, variables)

This code reads the specified variables from the data source and prepares them for processing.

To get the android_app_battery_level and android_app_step_count features, you can use the following code:

android_app_battery_level = datas.get_combined_data_by_variable('android_app_battery_level')
android_app_step_count = datas.get_combined_data_by_variable('android_app_step_count')

The data is now ready as a pandas DataFrame, and you can perform further analysis or visualization using standard pandas operations.

Step 4. Running a pipeline​

You can use compute_features method to configure the pipeline and compute features. Here is an example configuration:

input_config={
            "source_type": "local",
            "config": {
                "source_path": PATH
            },
            "data_format": "csv"
        }
feature_config={
            "location": "https://github.com/RADAR-base-Analytics/appdatacategorization",
            "feature_groups": ["AppCategorizationFeatures"],
            "feature_names": [["CategorizeApp"]]
        }
data = radarpipeline.compute_features(input_config, feature_config)
df = data['CategorizeApp']

This example can compute the app categorization features from the data source specified in input_config using the feature configuration provided in feature_config.

Show available pipelines​

To see the available pipelines, you can use the following code snippet:

from radarpipeline import radarpipeline
print(radarpipeline.get_available_pipelines())

The output will list all the pipelines that are available in the RADAR-pipeline module, allowing you to choose which one to run.

You can use the same function and methods in any Python script or application, making RADAR-pipeline a versatile tool for data processing and feature extraction in various environments.