[R] Help with Radius problem --update

Poling, William Po||ngW @end|ng |rom @etn@@com
Wed May 13 23:05:09 CEST 2020


Hello Abby and thank you for your response.

Your surly correct.

I have not worked a problem like this previously, however, I am learning fast.

I did not think I would need to apply mathematical formula-calculations for this task, math in general not my primary area of expertise, but always very curious.

However your point is provocative and now I am interested to learn more in that space as well.

By researching the literature on Radii, and disk partial coverage problem, and anything in stack overflow remotely relevant to my question I am slowly piecing a solution together.

What I described in the string is my best thinking in terms of delivering something, anything, at the moment that I can build on, refine, etc.

1. I have data set with 2352 geo locations in 41 states from Maine to California
2. I have determined, using density plot (thank to Jim L), that the most reasonable reference point in terms of realizing population density using distance among my 2353 is Brooklyn, NY.
3. Calculating distances from that point to the other 2351 based on a script referenced in stack over flow I simply plug in, manually-iteratively, distances (8 at the moment) and determine how many members are in that radius
4. Then determine the % decay from Max to Min of these 8 distances and settle on the one that has the highest concentration with the least decay
4. This piece (Item 3) I would like to make more efficient and was the point of my last submission to the R-Help list on this topic

For example: Radius2 = 1500 miles, Radius3 = 1200 miles, Radius4 = 1000 miles
The difference b/w Radius2 in terms of percent missing coverage and Radius3 is 0.057 while the percent missing coverage b/w Radius2 and Radius4 is 0.204

So my logic is that despite being a larger radius ,Radius3 vs Radius4, the trade off is more inclusive of the membership and therefore optimal among the 8 iterations. For the moment

I appreciate your response and any direction or advice would be most appreciated.

WHP

 



Proprietary

-----Original Message-----
From: Abby Spurdle <spurdle.a using gmail.com> 
Sent: Wednesday, May 13, 2020 3:17 PM
To: Poling, William <PolingW using aetna.com>
Cc: r-help using r-project.org
Subject: [EXTERNAL] Re: [R] Help with Radius problem --update

**** External Email - Use Caution ****

> "determine the largest concentration of members in the smallest radius"

I haven't read the whole thread, and I'm not familiar with this topic.
However, looking at it from an intuitive perspective, isn't the smallest radius zero.
If the concentration means the number of "members" divided by the area...
...then would the largest concentration be undefined, merely because of division by zero.

Perhaps your question is a valid one.
But if you're wanting a large pool of people to consider helping, I'd recommend a more (mathematically) precise definition of the problem you're trying to solve.

NOTICE TO RECIPIENT OF INFORMATION:
This e-mail may contain confidential or privileged information. If you think you have received this e-mail in error, please advise the sender by reply e-mail and then delete this e-mail immediately.  
This e-mail may also contain protected health information (PHI) with information about sensitive medical conditions, including, but not limited to, treatment for substance use disorders, behavioral health, HIV/AIDS, or pregnancy. This type of information may be protected by various federal and/or state laws which prohibit any further disclosure without the express written consent of the person to whom it pertains or as otherwise permitted by law. Any unauthorized further disclosure may be considered a violation of federal and/or state law. A general authorization for the release of medical or other information may NOT be sufficient consent for release of this type of information.
Thank you. Aetna


More information about the R-help mailing list