What dose the pred_results.predy mean?

[Jiale Ding]

Take the notebooks/GWR_prediction_example.ipynb as an example:

gwr_selector = Sel_BW(g_coords, g_y, g_X)
gwr_bw = gwr_selector.search(bw_min=2)
print(gwr_bw)
gwr_results = GWR(g_coords, g_y, g_X, gwr_bw).fit()

scale = gwr_results.scale
residuals = gwr_results.resid_response
pred_results = model.predict(pred_coords, pred_X, scale, residuals)

As suggest by the notebook and comment in source code, pred_results.predictions will output the predicted values at unsampled points.

However, I found that pred_results.predy is also callable, and the shape of output is same as the pred_results.predictions, but totally different in values.

It looks like the predy is the formal parameter of the constructor of GWR class. But it is also different with gwr_results.predy.

Though it doesn't affect normal use, I'm still curious as to where this predy came from.😂

[Taylor Oshan]

Predy is the in sample predicted values, also known as “y hat”.

[Jiale Ding]

Get it. Thank you. 👍