Comments on: OpenTemp.ParseStations() http://www.opentemp.org/main/2007/09/18/opentempparsestations/ An Open Analysis of the Historical Temperature Record Fri, 30 Jul 2010 01:31:21 +0000 http://wordpress.org/?v=2.2.3 By: John V http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-204 John V Fri, 30 Nov 2007 15:09:38 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-204 Thanks TCO. Thanks TCO.

]]> By: TCO http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-198 TCO Thu, 29 Nov 2007 23:55:10 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-198 practical statistics books: Check these out from the library (public libraries will do ILL, just ask them): * Box, Hunter, Hunter: Statistics for Experimenters (first edition is better written and more sharing of intuitions, second edition has more computer stuff) * Barker: Quality by Experimental Design Recommend also getting a time series book (not sure which one). Some people like Ross and Steve are heavy on time series but lacking on basic issues like confounding variables, interaction effects, hypothesis testing, type I/II errors, multiple regression, etc. (The same can also be true in reverse of course.) I think you are better off having a good DOE foundation to think about problems before getting wrapped up in autocorrelation and such. (and realize that DOE examines run order, so it is not ignorant of time series implications.) practical statistics books:

Check these out from the library (public libraries will do ILL, just ask them):

* Box, Hunter, Hunter: Statistics for Experimenters (first edition is better written and more sharing of intuitions, second edition has more computer stuff)
* Barker: Quality by Experimental Design

Recommend also getting a time series book (not sure which one). Some people like Ross and Steve are heavy on time series but lacking on basic issues like confounding variables, interaction effects, hypothesis testing, type I/II errors, multiple regression, etc. (The same can also be true in reverse of course.) I think you are better off having a good DOE foundation to think about problems before getting wrapped up in autocorrelation and such. (and realize that DOE examines run order, so it is not ignorant of time series implications.)

]]> By: John V http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-148 John V Wed, 21 Nov 2007 16:42:34 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-148 That sounds about right. You could actually get very close just by reducing the averaging distance using "/r=100" on the command-line and including only California stations. Cells that are not affected by any stations are excluded from the average temperature and the cells are pretty small (~50km on a side). That sounds about right.
You could actually get very close just by reducing the averaging distance using “/r=100″ on the command-line and including only California stations. Cells that are not affected by any stations are excluded from the average temperature and the cells are pretty small (~50km on a side).

]]> By: Clayton B. http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-140 Clayton B. Sun, 18 Nov 2007 22:44:52 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-140 Hey JohnV, With Anthony and crew nearly finishing up California - It'd be interesting to run Opentemp on only California. In order to to do this, one would need to a) change the boundary coordinates b) lower the averaging distance (100km may be ok?) c) reduce the cell size Is there anything I'm missing? Hey JohnV,

With Anthony and crew nearly finishing up California - It’d be interesting to run Opentemp on only California.

In order to to do this, one would need to
a) change the boundary coordinates
b) lower the averaging distance (100km may be ok?)
c) reduce the cell size

Is there anything I’m missing?

]]> By: John V http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-101 John V Mon, 05 Nov 2007 17:41:16 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-101 Thanks Clayton B, you're right. It's fixed for the next release (which might be a little while -- work got really busy). Thanks Clayton B, you’re right.
It’s fixed for the next release (which might be a little while — work got really busy).

]]> By: Clayton B. http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-93 Clayton B. Sun, 04 Nov 2007 21:05:10 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-93 I think the "## statations parsed" returns 1 too many. I think the “## statations parsed” returns 1 too many.

]]> By: John V http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-14 John V Thu, 20 Sep 2007 18:47:17 +0000 http://www.opentemp.org/main/2007/09/18/opentempparsestations/#comment-14 Should rename nLines to nStations and increment only if a station was successfully added. Should rename nLines to nStations and increment only if a station was successfully added.

]]>