This quickstart guides you through the installation of the
Google Cloud Pipeline Components (GCPC) SDK.

## Install latest release

Use the following command to install the Google Cloud SDK from the Python
Package Index (PyPI):

```python
pip install --upgrade google-cloud-pipeline-components
```

## Import a prebuilt component by using the Google Cloud SDK

After you install the Google Cloud SDK, you can use it to import a
prebuilt component.

For SDK reference information for supported components, see
the [Google Cloud SDK documentation](https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.23.0/api/v1/index.html).

For example, you can use the following code to import and use the
[Dataflow component](https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.23.0/api/v1/dataflow.html) in a pipeline.

```python
from google_cloud_pipeline_components.v1.dataflow import DataflowPythonJobOp
from kfp import dsl

@dsl.pipeline(
    name=PIPELINE_NAME,
    description='Dataflow launch python pipeline'
)
def pipeline(
    python_file_path:str = 'gs://ml-pipeline-playground/samples/dataflow/wc/wc.py',
    project_id:str = PROJECT_ID,
    location:str = LOCATION,
    staging_dir:str = PIPELINE_ROOT,
    requirements_file_path:str = 'gs://ml-pipeline-playground/samples/dataflow/wc/requirements.txt',
):
    dataflow_python_op = DataflowPythonJobOp(
        project=project_id,
        location=location,
        python_module_path=python_file_path,
        temp_location = staging_dir,
        requirements_file_path = requirements_file_path,
        args = ['--output', OUTPUT_FILE],
    )
```

## What's next

- Read the [Introduction to Google Cloud Pipeline Components](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/pipelines/components-introduction.md.txt).
- See all [tutorials that use the `google_cloud_pipeline_components`
  SDK](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/pipelines/notebooks).
- Get started with [Dataflow components](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/pipelines/dataflow-component.md.txt).