Is the reason why you think someone else plotted the graph because it rather comes off as amateur work by a pro or as professional work by an amateur ?
Anyway i wanted to finish all exercices to make sure i can put all the pieces together.
Im definitely ok with how to do it. Its easy.
But I cant understand the why if I cant systemize.
What you said makes more sense to me now.
Plz correct me if im wrong.
It starts with the previous chapter : it's impossible to get a straight line graph without an equation of the form y = ax + b. However some measurements even though they seem related by a law of propotionality can only be plotted in a non linear way (usually because they're false or too approximate). And because they re non linear their axis intercept and gradient cannot be deducted. i kinda understood that part before. What gave me a hard time is the replacement part. Well i got that part too now. It has to do with interpolation and extrapolation. not fake values. It obviously needs a linear graph to determine other values that have not been measured yet. So this replacement thing is used as a way of looking down to all values measured or not measured from a different angle where the whole thing appear as straight. Its pretty much like trying to determine earth distances if you looking to earth from the outer space. You cant. The curved form would mess your calculations. Youve got to take "straight" measurement from the ground - Sorry its the only metaphor i found- I got the idea but there are still missing pieces.
- can we extrapolate without using a straight line graph or is it the only way ?
- all the exercices i did refered to a particular equation, some of which were scientific, but what if im doing data analysis, and a law hasnt been established yet, and when established it is in no way relating to the forms seen in this chapter ?
y=ax^2 + b
y= a(1/x) + b
y/x = ax + b
y = ax^n
y = ab^x
y = ae^bx
These equations, is that it ? is linear regression only about those equations or should I expect more ?