By do data science, I mean the day to day stuff. So actually it's not that hard then if all they are doing is chaining together libraries.
Is there anything that this book would be missing the day to day stuff?
Is there anything that this book would be missing the day to day stuff?
But really understanding the failure modes of what you've made is much harder than that. Unfortunately I don't have a magic bullet or even a reference for that side of things. It's just something I've learned on the job. To paraphrase the Anna Karenina principle [0]:
All good models are alike; each bad model is bad in its own way.
Anyone who believes this is dangerous and shouldn't be allowed anywhere near a data science project. The programming is easy, but working with data and making good inferences is very hard.
But I understand what you mean, and completely agree that poseur amateurs who treat DS as just easy "number fluff" and expect fancy ML frameworks to solve all their problems are cancerous and ruin the reputation of DS as a field.
Yet the more I have worked with regular people at work, the more I have moved from the camp of always doing the "right thing" to hoping that people would just "do something". I can't fix every problem so maybe I'll just deal with the reality as it is and try to make best out of it. I won't start quizzing people at work about DS know-how but maybe then silently guide them towards understanding what they are doing instead of driving them out of the room and keeping them located somewhere far away from the data science team.