Is it better to use a string or wire to make a pendulem

and the heaver the weight does it make the pendulum more accurate

The wire would seem to be better than the string since in most cases for a given mass at the end the wire will not flex so much both due to the weight and the various states during the swing itself. The length is part of the equation determining the period, so having that vary isn't so good -- at least for repeatability between pendulums or precision of predicting their periods.

It should be obvious that there is some optimum weight, in between 0 weight and infinite weight, that works best for a given design and structure. So up to that point, a heavier weight is better.

A heavier weight of similar density will likely present a slightly larger cross section to the air it swings through meaning a slightly larger counter-force to the swing (which by itself isn't good) but it also will have larger sum of potential and kinetic energy as its energy reservoir (good because that will last longer with less percent lost each swing) so the net effect should be an improvement up to some point after which the wire or string will not fare so well. But this doesn't take into account the fixture at the top, where the weight presses against some kind of bearing system and adds it's own friction, too.

So again, everything is a balance. There will be an optimum weight that takes into account many factors, such as the way the pendulum is mounted, the strength of its structure and its own flexing, the string or wire or rod or whatever, and many other details if your goal is predictability and consistency of manufacture and cost and bulk and whatever factors are important to you at the time.