In my old system, I first make groups of three pairs of each two digits, plus a group of all that is left. Each number has a person, an action and an object attached to it, information I combine into a single image. The single digit is only an object.
12 = Inspector Lestrade - Knocking - Warrant
73 = Darth Sidious - Electrocuting - Death Star
70 = Darth Baras - Eating - Pizza
16 = Hannibal Smith - Smoking - Cigar
11 = John Watson - Shaving - Mustache
83 = Tom Cruise - Hanging - Cable
44 = Beast (X-men) - Brushing - Blue Fur
7 = Boomerang
Those I combine into three images, one for each of the sets of images I made.
12-73-70 = Inspector Lestrade electrocuting a pizza
16-11-83 = Hannibal Smith is shaving himself with a steel cable
44-7 = Beast has a few boomerangs stuck in his fur
Those images I place at my loci in whatever memory palace I want to use.
In my new system, which I am currently making, I do the same, but with three digit numbers instead of two, which allows me to end up with fewer images to remember. In this system I will keep the actions the same, as 1000 actions is really a lot. I also don’t have any objects or things for any two-digit numbers yet. So it would look like the following:
127 = Emperor Lrrr / (unspecified object)
37 = shattering
016 = Scrat / (unspecified object)
118 = Wendy (south park) / (unspecified object)
34 = spawning into creation
47 = Unspecified yet
So I would end up with two images, Emperor Lrrr shattering a yet unspecified object and Wendy spawning a yet unspecified object into creation.