carlosug
commented
2 months ago implemented at somef - header_level.py. It does the following:
- Extract Keywords from Headers: The extract_keywords_from_headers function identifies headers in the Markdown content and extracts keywords from them.
- Read JSON File: The read_json_file function reads the JSON file containing multiple README files.
- Extract README Contents: The extract_readme_contents function extracts the content of README files from the JSON data.
- Aggregate Keywords and Counts: We aggregate the keywords found in headers for each level (H1 to H6) and also count the occurrences of each header level.
- Plot Header Counts: The plot_header_counts function creates a bar plot showing the counts of headers by level. The modification made here assigns the x variable to hue and sets legend=False to suppress the legend, following the updated seaborn API requirements.
- Plot Heatmap for Each Header Level: For each header level, we determine the top 10 keywords and create a heatmap showing their distribution. We skip any header level that has no keywords.
level of hierarchy and how appears the themes including installation, usage, build, setup and running