DM41X PPC ROM Curve Fit

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

Trace: • Kenwood TM-D710 • Yaesu FT-891 • AOR AR7030 • Calculator Hardware • Becoming Teetotal • Radio • QRSS • Hi-Mound HK-708 Key • Yaesu FT-2980 • Icom IC-7200

You are here: Make It Tea » Public Area » Calculators » Hints & Tips » DM41X PPC ROM Curve Fit

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.

<-[[.:start]]


====== DM41X PPC ROM Curve Fit ======

The PPC ROM's ''CV'' function does similar Curve Fitting to the [[dm41x_curve_fitting|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 =====

<key>XEQ</key> <key>ALPHA</key>''CV''<key>ALPHA</key>

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 <key>SHIFT</key> <key>'E'</key> (Initialize)
  * ''X'' <key>ENTER</key> ''Y'' <key>'Σ+'</key>
  * ''X'' <key>ENTER</key> ''Y'' <key>'Σ+'</key>
  * ''X'' <key>ENTER</key> ''Y'' <key>'Σ+'</key>
  * ''X'' <key>ENTER</key> ''Y'' <key>'Σ+'</key>
  * etc.....

once the data is accumulated either

  * enter a fit-type digit (''J'')
    * 1 : Linear : '' Y = b*X + a ''
    * 2 : Exponential : '' Y = a*e<sup>bX</sup> ''
    * 3 : Logarithmic : '' Y = b*Ln(X) + a ''
    * 4 : Power : '' Y = a*X<sup>b</sup> ''
  * then press <key>'B'</key> to ''Solve Type J''

or

  * press <key>'E'</key> 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

<key>ENTER</key> ''Y<sub>1</sub>'' <key>'Σ+'</key>

<key>ENTER</key> ''Y<sub>2</sub>'' <key>'Σ+'</key>

<key>ENTER</key> ''Y<sub>3</sub>'' <key>'Σ+'</key>

<key>ENTER</key> ''Y<sub>4</sub>'' <key>'Σ+'</key>

...

and you accumulate as if you'd entered ''(1,Y<sub>1</sub>)'', ''(2,Y<sub>2</sub>)'', ''(3,Y<sub>3</sub>)'', ''(4,Y<sub>4</sub>)''... 

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 ===

^X^Y^
|1.1|5.2|
|4.5|12.6|
|8.0|20.0|
|10.0|23.0|
|15.6|34.0|

=== Entry procedure ===

  * <key>SHIFT</key> <key>'E'</key> ''Initialize''
  * ''1.1''<key>ENTER</key> ''5.2'' <key>'∑+'</key>
  * ''4.5''<key>ENTER</key> ''12.6'' <key>'∑+'</key>
  * ''8''<key>ENTER</key> ''20'' <key>'∑+'</key>
  * ''10''<key>ENTER</key> ''23'' <key>'∑+'</key>
  * ''15.6''<key>ENTER</key> ''34'' <key>'∑+'</key>

=== Fit to Linear ===

  * ''1'' (Linear Fit ''J'')
  *  Do fit : <key> 'B'</key>

=== Results ===

  * Z : '' 0.999035140  '' : r correlation coefficient
  * Y : '' 3.499147270  '' : a coefficient (y-intercept)
  * X : '' 1.972047542  '' : b coefficient (gradient)

'' Y = 1.972 X + 3.499 ''

== Estimations ===

Predict ''Y'' for ''X = 20 ''

  * ''20'' <key>'C'</key> 
    * '' 42.94009811 ''

Predict ''X'' for ''Y = 25 ''

  * ''25'' <key>'D'</key>
    * '' 10.90280649 ''



==== Best Fit Method ====

FIXME



===== Further Information =====


{{tag>calculator dm41x}}

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

Page upated : ~~LASTMOD~~

Navigation

Search

Toolbox