Chronology |
Current Month |
Current Thread |
Current Date |

[Year List] [Month List (current year)] | [Date Index] [Thread Index] | [Thread Prev] [Thread Next] | [Date Prev] [Date Next] |

*From*: bernard cleyet <bernard@cleyet.org>*Date*: Thu, 2 Apr 2020 13:33:31 -0700

If so, you did what colleges teach - giving undue weight to HIGH values and all but ignoring low values. But I am pleased to see someone is interested in tracking the time series.

That’s the impression I thought. Much earlier I fitted to early sections middle and end. Fortunately, I saved them, but didn’t think about the result. I’ll look and think. However, I think the fit (is a Marquardt) treats each datum equally unless one weights the data.

Have you been best-fitting log cases vs time by any chance?

Whether the scale (ordinate) is linear of log, the fit is the same.

bc delayed, soon

On 2020/Apr/01, at 21:27, brian whatcott <betwys1@sbcglobal.net> wrote:

Hmmmmm: your best fit of cumulative cases vs time puts most datapoints below the red line.

Have you been best-fitting log cases vs time by any chance?

If so, you did what colleges teach - giving undue weight to HIGH values and all but ignoring low values. But I am pleased to see someone is interested in tracking the time series.

Brian W

**Follow-Ups**:**Re: [Phys-L] Fwd: Re: six more days in Okla - least squares***From:*brian whatcott <betwys1@sbcglobal.net>

**[Phys-L] weighted linear regression using spreadsheets***From:*John Denker <jsd@av8n.com>

**References**:**[Phys-L] Fwd: Re: six more days in Okla***From:*brian whatcott <betwys1@sbcglobal.net>

**Re: [Phys-L] Fwd: Re: six more days in Okla***From:*bernard cleyet <bernard@cleyet.org>

**Re: [Phys-L] Fwd: Re: six more days in Okla***From:*brian whatcott <betwys1@sbcglobal.net>

- Prev by Date:
**Re: [Phys-L] modeling is hard** - Next by Date:
**Re: [Phys-L] modeling is hard** - Previous by thread:
**Re: [Phys-L] Fwd: Re: six more days in Okla** - Next by thread:
**Re: [Phys-L] Fwd: Re: six more days in Okla - least squares** - Index(es):