# Debug transforms

The Palantir extension for Visual Studio Code supports inspecting variables when debugging Python transforms. Complex data structures such as DataFrames can be loaded into the extension's **Preview** panel, allowing you to verify transform accuracy and identify issues before [initializing a build](../transforms-build/index.md).

## Start a debug session

To use debugging features with the Palantir extension for Visual Studio Code, open a Python file containing a transform. Set at least one breakpoint in your code to pause execution when that code is executed. When the breakpoint is reached, Visual Studio Code's **Debug** panel will open automatically and display the variables currently in scope.

To start a debug session, select **Debug** above the transform.

![The "Debug" option above an example Python transform.](../../../assets/eeeb7e243b56b66b9f170794986a0d06c496d525b2eabe55d3e7ab89a58fe858.webp)

When a breakpoint is reached in your debugging session, you can inspect supported variables using the **Preview** panel through the following two methods:

* Select **🔎 Inspect current value of …** above the variable you wish to inspect

![The "Inspect variable" option above an example Python transform.](../../../assets/620b92401f5cbcc686501d4016e3613ad20b97013f9f50e23f67d240cdeb159a.webp)

* Right-click on the variable in Visual Studio Code's **Debug** panel and select **Palantir: Inspect variable**.

![The "Inspect variable" option in Visual Studio Code's Debug panel.](../../../assets/8b2143f951aeb25279959f7ad76c189a5a63b198c5d192bb86a103cda4a9b95e.webp)

Inspecting a variable will load it into the extension's **Preview** panel. You can then view the variable's contents and use the **Preview** panel's features to explore the data, such as viewing column statistics.

![A variable being inspected through the extension's Preview panel.](../../../assets/9f745c44a9195d4c8817bdcdeeae27680b10484dd52e6852e45ed03960b7515d.webp)