User talk:Dcromley

From the first sentence: 180 grains converts to 11.6638 grams. 3/8 troy ounces converts to 11.6638 grams. I've spent a lot of time trying to find out where "(11.3398038 grams)" comes from. This needs to be justified, at least in the talk. Dave C (talk) 03:20, 8 February 2019 (UTC)[reply]

Your Collatz tree

Hi, I had a look at your tree at http://dbarc.net/yr2024/collatz1.0.pdf

I do like your idea of showing /2 on top, (*3+1)/2 at the bottom. It is simple and elegant.

Four suggestions:

Add the 1->2 to the bottom part, the cherry on the cake, the final loop.
Try half ovals in stead of straight lines, with the steep bit at the end point. This will indicate direction in a subtle way.
Omit the small circles as when using half ovals, you will see: 2 endings, 1 start at 2, 5, 8, 11, ...
Submit your graphic at: commons:Special:UploadWizard and post it on the Collatz Conjecture talk page.

Uwappa (talk) 17:46, 31 July 2024 (UTC)[reply]

Your response was very well received (by me). Then I lost track of it, and was mortified. But now I FOUND it! Very good input! I (and my grandson) realize I/we got overly enthused (are crackpots), but have enjoyed it. The basic fallacy was pointed out, and I commit to being more meticulous in the future. We still have fun at it with no expectations. Please see http://dbarc.net/yr2024/uwappa1.html, which is just one of programs that I plan to post. (It's not complete.) Your suggestion #1 is troublesome. I hope to communicate with you again. Dave C (talk) 23:18, 16 August 2024 (UTC)[reply]

Thank you, a HTML page named after me, wow that is a first, thank you! You and your grandson must have enjoyed yourself programming those nifty graphics!

The branching level at the bottom confused me at first, 1, 5, 2, 4, 3, 5, 4. Huh? It took me some time those are the number of steps to the final one. Having two lines with number looks a bit cluttered. A possible alternative: replace the branch numbers by a vertical position, branching level 0 at the top, level 1 pixel lower, etc.

I did try out my own suggestions using a simple drawing program. The resulting graphic was 'clean', clear, no clutter. But it was just a quick try-out, not good enough to upload and share.

I hope you and your grandson can find the time to program these suggestions and create a professional looking graph that is worth uploading.

What is the problem with my first suggestion? 1 is just another odd number, isn't it? Uwappa (talk) 00:36, 17 August 2024 (UTC)[reply]

Your 'clean' graphic has my interest.

RE: first suggestion -> The problem is the 1 2 1 2 .. loop. It was easier for me to just eliminate the 2-1 branch.

It takes me a long time to come up with something presentable, so it may be awhile before you hear from me.

And I have very little experience within wikipedia (I am a $ contributer).

Thanks again. Dave C (talk) 01:53, 17 August 2024 (UTC)[reply]

No worries, take your time. The Collatz conjecture has been puzzling mathematicians for decades so there is no rush.

I do like your idea of having a program that draws the graph. This allows you to change the program, experiment with different graphs styles.

Some additional ideas for a clean graph:

Use the vertical dimension for (3x+1)/2 only. Let /2 run horizontal.
Show branching level with shades of grey, e.g. black for level 1 and very light grey for the highest branch level. This will visually connect lines, showing direction, going from light grey to black as they approach 1.
Alternative: use line thickness for branch level, e.g. very thick for the short line at level 1, very thin for high levels. Change line thickness only at knots where 3 lines meet. Two thin lines will 'merge' to a thicker line. This will 'show direction'.

Success! Uwappa (talk) 06:55, 17 August 2024 (UTC)[reply]