Math programs used

Table 8: Math programs used.

#	Which spreadsheet - data base - math programs do you use or prefer to use to do ``meta-analysis'' of results?	use?		prefer?
		Number Yes	Percent Yes	Number Yes	Percent Yes
32	Excel	18	28.1	5	7.8
34	Origin	20	31.3	13	20.3
36	Kaleidagraph	9	14.1	5	7.8
38	SigmaPlot	8	12.5	5	7.8
40	Jess Brewer's DB	23	35.9	10	15.6
42	TRIUMF Physica	18	28.1	11	17.2
44	PAW	4	6.3	3	4.7
46	Mathematica	16	25.0	3	4.7
48	Matlab	7	10.9	3	4.7
50	Maple	12	18.8	1	1.6
52	IDL	0	0.0	0	0.0
54	Other: Total	15	23.4	9	14.1
56	Other: own	5	7.8	2	3.1
58	Other: Wimda	3	4.7	3	4.7
60	Other: Gnuplot	3	4.7	0	0.0
62	Other: Grapher	2	3.1	1	1.6
64	Other: Derive	1	1.6	1	1.6
66	Other: Java GUI	1	1.6	1	1.6
68	Other: (XM)Grace	1	1.6	1	1.6
70	Other: Igor	1	1.6	1	1.6
72	Other: GLE	1	1.6	1	1.6

Table 9: Which math programs used and preferred, sorted by category.

	mean per respondent
	Used	Preferred
Total	2.3 $\pm$ 1.2	1.1 $\pm$ 0.9
Spreadsheet	0.9 $\pm$ 0.9	0.4 $\pm$ 0.6
Non-spreadsheet	1.3 $\pm$ 1.2	0.5 $\pm$ 0.7
Commercial	1.4 $\pm$ 1.2	0.6 $\pm$ 0.7
Free	0.7 $\pm$ 0.8	0.4 $\pm$ 0.6

Users were asked which spreadsheet - database - math programs they use to summarize (e.g. do meta-analysis) of their results (Table 8). Which kinds of programs people use indicates which formats of results output muSR analysis programs should produce. A degree of flexibility would be useful, such as being able to choose between tab-delimited columns and DB format.

This question also served to highlight several interesting issues:

• As needs dictate, people use the style of program best suited for the task at hand. People use a mean of 2.3 $\pm$ 1.2 math programs.
• Of the spreadsheet programs (Excel, Origin, Kaleidagraph, SigmaPlot), Origin was the favorite. There were no free spreadsheets used.
• Of the non-spreadsheet programs (DB, Physica, PAW, mathematica, matlab, maple and IDL), only DB, Physica and PAW are free. Matlab is available in a freeware version called Octave. DB and Physica were the favorites.
• People use one hundred percent more commercial programs than free programs and prefer fifty percent more commercial programs than free programs (see Table 9).
• People use fifty percent more non-spreadsheet programs (see Table 9), while the preference is almost evenly slit between spreadsheet and non-spreadsheet programs.
• These differences appear to be cultural in that those who prefer to use free software tend to be anti-Microsoft, and tend to prefer direct control over software and data formats of results, etc.. Some use of free software may be due to financial considerations, although no one mentioned this specifically. Those who use more commercial software, especially spreadsheets, are probably less inclined to do conventional programming.
• Three respondents wrote-in the new analysis program Wimda, which appears to be a strong contender to both DB and msrfit.
• Facilities should probably ensure that users on site have access to both a common spreadsheet program and to a command-based matrix-based math program.