Let's start from MODL which is a “sub-sub-menu”. It has LINF, LOG, EXPF, PWRF and BEST.

If the LINF is active the calculator will try to fit the data in a y=Mx+B model. If the LOG is active the calculator will try to fit the data in a y=M lnx+B model. If EXPF is active the model is y=B exp{Mx} and PWRF is y=B x^M.

How can I select the model? This depends one what kind of data you are using. If you really don't know try the BEST function which will change to what can be the best model.

The other functions in CFIT sub-menu are quite simple.

YINT: Gives the value of B (the name comes from LINF model which is of course the main model).

SLOPE: Gives the value of M (again this comes from LINF model). FCSTX: Just run the model for

a given y and returns a estimated x. FCSTY: Returns a estimated y for a given x. CORR: Returns a number between -1 and 1 (the correlation coefficient) which tells us how good is the fit. 1 is the best result.

Of course if you have only two data the calculator will find LINF to best model and the correlation coefficient will be 1, which does not mean the estimation will be good!

9.2 The second line: ALLΣ, LINΣ, ΣREG and ΣRG? Functions.

Unlike HP-33S which has special variables for statistic, the HP-42S uses the normal Rnn memories from R11 to R23 in the following way.

R11 for Σx, R12 for Σx2, R13 for Σy, R14 for Σy2, R15 for Σxy and R16 for n.

This is like HP-41 and if LINEΣ is active it is all we have. If ALLΣ is active we have also

R17 for Σln x, R18 for Σ(ln x)2, R19 for Σ ln y, R20 for Σ(ln y)2, R21 for Σln x ln y, R22 for Σ x ln y and R23 for Σ y ln x.

Why should I know this? Because now we know how to access statistical data in programming and also we know we cannot put important data in memories from R11 to R23 if we are going to use statistics. (33S is better)

Of course unless you need HP-41 compatibility keep the calculator always in ALLΣ mode.

We said the statistical data start in R11 but you can change this using ΣREG and to view where it is starting just use ΣRG? Function.

Conclusion: Four useless functions!:(