ceiling rose has 4 cables coming in.. cable one has red, yellow, blue. second has red, black. third has red, black, earth. forth has red, black, earth. light switch is a double switch one for concealed lighting, second is for main light.
Are we to presume switching is only single way, there are no other switches for the same lights elsewhere? You need to buy yourself a multimeter so you can test this properly rather than just guessing at where to put wires and seeing what happens, as that is a dangerous way to work.
My guess would be that the 3 core red/yellow/blue runs to the switch, with red as the common and yellow/blue as switched live returns. One red/black will be incoming live, another the outgoing supply, and the final red/black probably feeds the concealed lighting.
With a meter and all connections isolated in a terminal block then it shouldn't be too hard to determine what everything does. Using as a voltmeter with all cores terminated into a terminal block and the supply on you'll be able to identify the incoming supply as there will be 230v between black and red. With supply off, you can use the meter as a continuity checked to identify the switched lives.
Again, with the supply off, you could do a continuity check to another ceiling rose on the same circuit using a wander lead and identify the outgoing live. As a final check (again, supply off) you could check for resistance across the remaining unidentified cable between the black and red cores, if the concealed lighting is mains fed (no transformer) then you could expect a
relatively low resistance. That said, a more reliable test would be to con check back to the fitting(s) themselves, although if my theory is right then the concealed lights would be fed off the red/black cable connected to the blue core.
I would have expected all these cables to have a CPC, so I'd check again and be absolutely sure that the bare copper earth conductor hasn't been chopped back.
If you're not confident with any of the above then call an electrician.