T_wide used in calculations

[ajj]

Check whether use of T_wide in calculations for normalization is correct.

This is a problem when data isn't measured out far enough in Q.

[srkline]

T_Wide should be normalized by the monitor count, but it isn't.

The T_wide value is found from a running average of angles > 2.0

Do we need intervention if this is not present say manually enter a value, or use the "lowest" 5 points available, and display a warning message that T_wide may be suspect. It's clear from a plot of the raw data if the transmission detector counts have leveled off, so suggest that the users look at that.

[srkline]

don't just normalize - either look for and ignore zeros, or normalize the sum of trans counts by the sum of monitor counts

[srkline]

now normalized to the monitor counts. Also, when data > 2 degrees is not present, a warning is presented and the last 4 data points of whatever is present is used (and appears to be not too bad).

r540