changes in ralg and derivatives check

Some major changes for NLP/NSP ralg solver.

I have committed some changes in check user-supplied derivatives.

Old-style difference is not a good choice because sometimes you have values to compare like 1.24*10⁵ (user-supplied) and 1.23*10⁵ (finite-difference), while sometimes 1.24*10^-5 (user-supplied) and 1.23*10^-5 (finite-difference).

The new way is based on comparing
1 - derivative_user/derivative_finite_diff
to maxViolation value

The parameter maxViolation can be used as prob field or via for example p.checkdc(maxViolation=0.05)
Note that x remains to be 1st argument for checkdc, checkdf, checkdh (default p.x0), and can be used as kwarg as well: p.checkdh(maxViolation=0.05, x = 0.5*(p.lb+p.ub)), that is same to p.checkdh(0.5*(p.lb+p.ub), maxViolation=0.05).

So new text output is like this one below (this one is produced by updated examples/checkDerivatives.py file).

Note that RD (relative difference) is defined as

int(ceil(log10(abs(Diff) / maxViolation + 1e-150)))

where

Diff = 1 - (info_user+1e-150)/(info_numerical + 1e-150)

(those small numbers are used to suppress zeros)
################################

OpenOpt checks user-supplied gradient df (shape: (30,) )
according to prob.diffInt = [9.9999999999999995e-08]
lines with 1 - info_user/info_numerical greater than maxViolation = 0.01 will be shown
df num         user-supplied     numerical               RD
    0             +7.000e+00     -8.000e+00              3
    8             -2.291e+00     -1.029e+01              2
max(abs(df_user - df_numerical)) = 14.9999995251
(is registered in df number 0)
========================
OpenOpt checks user-supplied gradient dc (shape: (2, 30) )
according to prob.diffInt = [9.9999999999999995e-08]
lines with 1 - info_user/info_numerical greater than maxViolation = 0.01 will be shown
dc num   i,j:dc[i]/dx[j]   user-supplied     numerical               RD
    32             1 / 2         +1.417e+01     -8.323e-01              4
max(abs(dc_user - dc_numerical)) = 14.9999999032
(is registered in dc number 32)
========================
OpenOpt checks user-supplied gradient dh (shape: (2, 30) )
according to prob.diffInt = [9.9999999999999995e-08]
lines with 1 - info_user/info_numerical greater than maxViolation = 0.01 will be shown
dh num   i,j:dh[i]/dx[j]   user-supplied     numerical               RD
    58            1 / 28         -4.474e+01     -5.974e+01              2
max(abs(dh_user - dh_numerical)) = 14.9999962441
(is registered in dh number 58)
========================