Hi all, As happens from time to time, discussions on this list appear regarding the use of popular spreadsheets for statistical analysis. One such thread (post of mine) is here: http://maths.newcastle.edu.au/~rking/R/help/03a/6326.html While not advocating such use, these discussions have referenced articles that provide independent reviews of these applications and issues of accuracy, etc. Based upon a posting to Linux Today (http://linuxtoday.com/news/2004022600226OSBZGN), I became aware this evening of a new paper by the well known B.D. McCullough, who has published many such reviews of both spreadsheets and other statistical applications. In the latest paper (pdf available from http://www.csdassn.org/software_reports.html), Prof. McCullough reviews Gnumeric as an example of how an open source project has responded to prior criticism of issues, while raising well known issues with Excel, at least through XP. He leaves open to further research, any improvements to Excel 2003. One lingering criticism is Gnumeric's RNG, however Prof. McCullough (who references Prof. Ripley's 1990 PRNG paper) indicates in this article that the Gnumeric team has incorporated the Mersenne Twister to a beta version of Gnumeric, which may already have been released by now. This modification ameliorates this concern. For those wishing to stay abreast of such issues, this new paper may be of interest. It puts the open source efforts of the Gnumeric team (http://www.gnome.org/projects/gnumeric/) in a very favorable light. Best regards, Marc Schwartz
Gnumeric - 1 Excel - ?
2 messages · Marc Schwartz, A.J. Rossini
Marc Schwartz <MSchwartz at medanalytics.com> writes:
For those wishing to stay abreast of such issues, this new paper may be of interest. It puts the open source efforts of the Gnumeric team (http://www.gnome.org/projects/gnumeric/) in a very favorable light.
Thanks. For those using gnumeric to munge excel data, be careful. Very careful. Just found a spreadsheet with "missing" columns (exist in excel, and do not display in gnumeric). If I'd known they were there, I'd not have had to recopy from an older dataset. Realigning/merging at 4am isn't fun. Alternatively, if I had exported first into R (via CSV), I'd have found them. Not sure what the lesson is, but at 5am, I'm just not too sure about anything except that I've now got a clean and validated dataset. best, -tony
rossini at u.washington.edu http://www.analytics.washington.edu/ Biomedical and Health Informatics University of Washington Biostatistics, SCHARP/HVTN Fred Hutchinson Cancer Research Center UW (Tu/Th/F): 206-616-7630 FAX=206-543-3461 | Voicemail is unreliable FHCRC (M/W): 206-667-7025 FAX=206-667-4812 | use Email CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{dropped}}