Estimate Food Network Demo

This demo showcases how to estimate a semantic network from fluency data in the “foods” category using the Conceptual Network method. The script walks through data cleaning, error detection, and network construction based on co-occurrence logic.

—

Steps Performed:

  1. Load fluency data grouped by participant and apply spell correction.

  2. Detect and list: - Intrusions: Items not present in the category scheme - Perseverations: Repeated items within the same list

  3. Flatten the data (convert hierarchical data to list-level) to prepare for network estimation.

  4. Generate a Conceptual Network using item co-occurrence.

  5. Export the resulting semantic network as an edge list (foods_network.csv).

—

Key Parameters (in `Fitinfo`):

  • cn_windowsize: Size of the sliding window for co-occurrence

  • cn_threshold: Minimum list frequency required for an item to be included

  • cn_alpha: Significance level for co-occurrence edge inclusion

—

Relevant Functions Used:

  • snafu.load_fluency_data

  • snafu.intrusions / intrusionsList

  • snafu.perseverations / perseverationsList

  • snafu.Fitinfo

  • snafu.conceptualNetwork

  • snafu.write_graph

Output:

  • demos_data/foods_network.csv — A full edge list of the estimated semantic network.