Large p-values incorrectly approximated

[avsmith]

In the file Algorithm.cpp, p-values for T less than 1.6 are being approximated by the following (likely for speed reasons):

                if( t < 1.6)
                {
                    t1=4.4-t;
                    ptemp=0.5-0.1*t1*t;
                }
                else
                {
                    ptemp=erfc(t*one_oversqrt2);
                }

However, the approximated p-values for t<1.6 are too high. Plotting p vs t results in a discontinuity at 1.6.

[lckarssen]

Thanks for reporting this and providing a fix! I'll have a look at it next week, trying to see if I can find impact on speed.