However, the film is Indian and I haven't actually seen it. I recall an Indian math historian speaking about this film at our university colloquium several years ago. He boasted, "Ben Kingsley was interested in depicting Ramanujan but was turned down for the role because he was only half Indian".
After trying a few different iterations of the model with different distributions, it really seems like the soap-film smoother's boundary constraint is setting the partial effect of the smoother to zero at the boundary, rather than forcing the response variable to zero at the boundary. Is my understanding correct?
Issue with boundary condition in a GAM with a soap film smoother
I conducted a research with 90 participants who were exposed to a film stimuli for 60 minutes and their physiological responses were measured continuously. I would like to test the causal relations...