I want to smooth out the general noise so i will, but there is an around 10second sine wave along all of the data. When there are spikes of force this just gets added to the sine wave, as if it is a baseline. therefore i would just like to remove the sine wave and bring it back to the data it should be .

That is incredibly difficult to do, it is just down to the maths, not a node-red problem. If you know exactly the frequency of the signal you might be able to us some sort of notch filter, but I don't hold out much hope of success.
Is the frequency of the sine wave known exactly, and constant? Is it of constant amplitude? If so then you might be able to use a generated compensation signal to subtract from the data. Alternatively you could record significant periods and then analyse the data retrospectively to determine the sine wave parameters and then subtract it. It is a pretty complex data analyses problem, that is above my maths knowledge. I think you might have to consult an expert in such things if this is critical to you. As I said, the problem isn't how to do it in node-red, but how to do it at all.