Try to learn something about everything, and everything about somethingThomas Huxley “Darwin's bulldog” (1824-1895)

This is an old revision of the document!


Hints & Tips

DM41X PPC ROM Curve Fit

The PPC ROM's CV function does similar Curve Fitting to the Advantage Pac CFIT program, without the limitation on negative data point values.

Negative X values are acceptable for Linear and Exponential curves

Negative Y values are acceptable for Linear and Logarithmic

If any data point has BOTH X and Y negative then only Linear is possible.

Running CV

XEQ ALPHACVALPHA

Unlike CFIT there is no displayed menu for the top-row buttons, so you need to remember the functions.

Key A B C D E
Shift ∑- Initialize
Σ+ Solve Type J Y X Solve Best Type

Unlike normal Stats entry (where Y-value is entered first, then ENTER, then X-value is entered second, so that the X-value is in the X register before pressing ∑+) here we enter in “conventional” order.

  • Clear old data SHIFT E (Initialize)
  • X ENTER Y Σ+
  • X ENTER Y Σ+
  • X ENTER Y Σ+
  • X ENTER Y Σ+
  • etc…..

once the data is accumulated either

  • enter a fit-type digit (J)
    • 1 : Linear : Y = b*X + a
    • 2 : Exponential : Y = a*ebX
    • 3 : Logarithmic : Y = b*Ln(X) + a
    • 4 : Power : Y = a*Xb
  • then press B to Solve Type J

or

  • press E to Solve Best Type

The solution appears as

Z : r correlation coefficient Y : a coefficient X : b coefficient

You can now get estimates of Y (for a given X) or X (for a given Y)

You can accumulate more data (to add to what's already there) and do another FIT calculation.

If you want equally spaced X coordinates for single variable Y-data CV puts incrementing digit in X so you simply

ENTER Y1 Σ+

ENTER Y2 Σ+

ENTER Y3 Σ+

ENTER Y4 Σ+

and you accumulate as if you'd entered (1,Y1), (2,Y2), (3,Y3), (4,Y4)

If you have X and Y data then the automatically incrementing X is overwritten by the manually entered real X values.

Examples

Manual Linear Fit

Data

Negative X and Y values allowed for a Linear Fit

XY
-1.5 -3
1.15.2
4.512.6
8.020.0
10.023.0
15.634.0

Entry procedure

  • SHIFT E Initialize
  • 1.5CHS ENTER 3CHS ∑+
  • 1.1ENTER 5.2 ∑+
  • 4.5ENTER 12.6 ∑+
  • 8ENTER 20 ∑+
  • 10ENTER 23 ∑+
  • 15.6ENTER 34 ∑+

Fit to Linear

  • 1 (for Linear Fit J = 1) B

Results

  • Z : 0.995734680 : r correlation coefficient
  • Y : 2.017167182 : a coefficient (y-intercept)
  • X : 2.113978698 : b coefficient (gradient)

Linear Equation : Y = 2.114 X + 2.017

Estimations

Predict Y for X = 20

  • 20 C
    • 44.9674114

Predict X for Y = 25

  • 25 D
    • 10.87183747

Best Fit Method

FIXME

Further Information

See also : DM41X PPC ROM Solve & Integrate

Page created : 16/02/26 20:13 GMT

Page upated : 17/02/26 07:53 GMT


Navigation