Lookup dissimilar word describer data

Hello,

I'm sorry but this is a bit confusing. Can you be more detailed about the "word describer blocks", and if you have to do it manually, how would you do it?

And also, can you show your expected results in your sample file? Looking at your attached file was really confusing when I don't know where to start - should I compare and pick the differences between each encoded image as a whole, or just find ten words from all 100 images that has the least frequency, etc.

Hi Lemice,
Yes you should compare and pick the differences between each encoded image as a whole.
First one is randomly picked and each of the 10 is different from each other as a whole.

The "word describer blocks", are a 30x10 cell array which content words to express a image as a whole.
Its hard to do it by hand as this is the result I'm using excel... but as a simple example as follows:

red, round, many

most difference should have different characteristic in each block of 10 thus the next one is:

blue, square, flat,

you any need to pick the different of words not any group of characteristics.
keep in mind same block are similar so you just can't pick random 10 out of 100 is the biggest distance each work block has from one another. hope that more clear to you...

blue, square, flat,
red, round, many
green, boxy, wet

etc...

To be honest, I have no idea what you are trying to do ...

Excel is a tool after all, and you have to give it instruction on how to do a things. It cannot solve anything on its own.

Here is a attached file of what I imagine I would do.

To use it, just change the number of image you want to check on B7 to anything from 1 to 100. The following 9 images will change itself. Right next to the number of each image you will see the number of similar words of the image with the first image you chose.

For instance, in this attached file, the chosen image is 1 (B7 = 1), the 2nd image is 57 with only 15 words similarly to the first, the 3rd image is 64 with 16 similar words to the first, and so on.

I also put the number of similar words result next to a hundred images on your list (to make it easier for the formula to find, and to avoid giant formula with unnecessary repeated step during calculation).

Lem, I thank you for you time.

You got it all correctly working with one last rule; which is the hardest to explain
I need these 9 'work blocks' to not only have different words from the first word block 1,
but also have different words (as close as possible) from each other too...

This is needed as the 9 images have similar words with each other, I needs the 9 images to have different words from each other.
could this be done you are so close...

I suggest you check out the VBA forum, because you will not like what I am going to tell you.

After you choose a picture, there are 99 left. In order to choose the second one, I will need to do 99 calculations to find the most different image comparing to the first. For the third one, that will be 98*2 calculation because I have to compare the rest 98 images with number one and two, then find the average of them. For the fourth one, it will be 97*3, and so on. The last one will be 91*9. The total number of step to calculate will be huge, and just from the sample file above, the calculation is only one step and it is pretty slow already.

Unless you have a more specific rule to cross out and choose more image, I'm afraid this will not a good way to continue.

Please do not create another thread in the vba forum for this question. If you would like to obtain a solution through vba, please report this thread to the mods by click on the triangle icon with the exclamation mark. One of the mods / admin will move this thread for you.