I used TrigReduce on products of cosines in Mathematica.
It led to me to develop the formula $\prod_{i=1}^k\cos\left(\beta_i\right)=\sum_{j=1}^{2^{k-1}}\frac{\cos\left(P_j\right)}{2^{k-1}}, k \geq1$ where $P_j$ is a permutation of the signed sums of $\beta_i$'s, i.e. $\beta_1+\beta_2+\beta_3+\beta_4+\beta_6$ and $\beta_1-\beta_2+\beta_3-\beta_4+\beta_6$ are two of the 32 permutations with $k=6 $ ($\beta_1$ always has a positive coefficient).
How would I prove this? Is this formula out there in literature?
I tried proving it by using the chain rule and then reverse because integration is linear over sums, and derivatives make sine functions more "summy". Then was going to use a recursive formula to prove by induction. But that line of reasoning doesn't explain how the $\beta$'s get in the same cosinusoidal term. Also, nothing in my undergrad used this and I couldn't find anything on google. And Mathematica doesn't give the general formula when I use k instead of a fixed number. ( I also get something similar with sinusoidal products).