Moving forward, all updates and results shared must adhere to the following guidelines.

These are non-negotiable and essential for maintaining clarity, professionalism, and effective communication.

1. Provide Context and Details

When sending updates about something that worked or didn’t work, do not simply state the outcome.

Include:

• What you did: Briefly explain the steps or methods you followed.
• Why it worked/didn’t work: Provide reasoning, evidence, or error messages to support your claims.
• Next steps: Clearly outline what you plan to do next to address the issue or build on the success.

Example of an unacceptable update:

"The model didn’t work."

Example of an acceptable update:

"The model failed to provide a good output likely due to a domain shift in the testing data. I think we should fine-tune the model on dataset-X in order to obtain better results for our usecase"

2. Showcase Your Results Properly

Simply sending raw files (e.g., shapefiles, GeoJSONs, CSVs) is Not Enough. You must include visualizations and analysis to make your results meaningful.

• Overlay results on maps: Use tools like QGIS or Python libraries (e.g., Folium, Geopandas, Matplotlib.. ) to visualize your results spatially.
• Include plots: Provide charts, graphs, or heatmaps to highlight trends, patterns, or anomalies.
• Draw conclusions: Assess whether the results are good or bad. Highlight areas of success and identify potential issues or limitations.

some helpful libraries :

• Contextily : https://contextily.readthedocs.io/en/latest/intro_guide.html
• Geopandas : https://geopandas.org/en/stable/