Adding a Custom Data Loader

This guide explains how to integrate a new data loading method into the application, in addition to the built-in CSV and HDF5 readers.

1. When to Customize

You may want to implement a custom loader when dealing with:

  • JSON files
  • Excel spreadsheets
  • Proprietary binary formats
  • Direct connections to databases or instruments

2. Where to Modify

The DataLoader class handles all file loading operations and is located in:

gcfpy/utils/data_loader.py

3. Implementing a New Loader

Step 1 — Add a Method

Define a new method in the DataLoader class:

def load_json_format(self, file_path):
    try:
        import json
        with open(file_path, "r") as f:
            data = json.load(f)
        df = pd.DataFrame({"X": data["X"], "Y": data["Y"]})
        return df
    except Exception as e:
        self.show_error_message(f"Failed to load JSON: {e}")
        return None

Make sure your method returns a pandas.DataFrame with columns X and Y (and optionally Z, X_err, Y_err).

Step 2 — Register the Extension

Update the main read_file() method to detect and dispatch based on the file extension:

if file_name.endswith(".json"):
    self.df = self.load_json_format(file_name)

You may also update the file dialog to list .json as a supported extension if needed.

4. No Further Changes Needed

Once your custom loader returns a valid DataFrame:

  • The X/Y table will be populated
  • The data will appear in the plot
  • The fit panel and toolbar will be enabled

No need to modify any part of the fitting or visualization pipeline.