Documentos de Académico
Documentos de Profesional
Documentos de Cultura
CH5 Data Acquisition and Representation
CH5 Data Acquisition and Representation
Use formal data sheets or notebooks to record all data and other
observations.
General Guidelines When Conducting Experiment
Calibrate all your measuring instruments for its validity before you start
using them. Read or review equipment manuals for instrument accuracy
and precision.
Have your data verified before leaving the laboratory. Often your instructor
or lab assistant will be able to help you confirming your results. Some data
sheet even requires an endorsement from the respective authority.
Format of formal report
1. Objective Section-state aim and objectives clearly
2. Equipment Section-list equipment with sketches and diagrams duly labeled
with figure number
3. Theory Section- brief mention of theory or related equations
4. Procedure or Methodology Section-write in steps how the experiment is
conducted
5. Data Section-properly rounded and use appropriate significant figures
6. Calculations Section-sample calculation, error analysis in logical sequence
7. Results Section-tabular, graphical or combination of both forms
8. Conclusion Section-achievement related to objectives, practical
application, validity, errors encountered, suggestions for improvement
Suggestion format for this course
1. Introduction
2. Objectives of the Experiment
3. Equipment /Apparatus
4. Theory
5. Procedure
6. Data and Sample Calculation
7. Results and Discussions of Results
8. Sample Calculations
9. Appendices
Empirical function are generally defined as those based on values from
experiment and observation rather than theory.
Mathematical expressions can be modeled to fit experimental functions
Types of empirical functions :
Linear : y = mx + c :
• Linear function will plot as a straight line.
Exponential: y = a e mx
• When plotted on semilog paper will be linear
• In log form log y=mx log e + log a
Power : y = ax m
• In log form log y=m log x + log a
Periodic : y = sin Ө
• repeats its values in regular intervals or periods. The most important
examples are the trigonometric functions, which repeat over intervals of
length 2π.
• are used throughout science to describe oscillations, waves, and other
phenomena that exhibit periodicity
Different methods or techniques are available to arrive at the “best straight
line” fit.
Three methods commonly employed for finding the best fit are as below:
1. Selected points method
2. Average method-line location is positioned to make algebraic sum of
absolute values of the differences between observed and calculated
values of the ordinate equal to 0
3. Least square method
Selected points method
• Determine equation that best fit data that exhibit linear relationship.
( y = mx + c )
• Select line that goes through as many data points as possible and has
approximately the same no of data points on either side of line.
• Select two points on the line at a reasonable distance apart:
• let’s say A(x1, y1) and B(x2, y2), the slope m is determined as
y2 − y1
m= (1)
x2 − x1
• the y-intercept c may be either obtained directly from the graph or by
substituting point A or point B back into the equation. Rearranging the
equation y = mx + c in the form c = y – mx and using point A(x1, y1) for
example, we have
c = y1 – mx1 (2)
Selected points method
Least square method
• Known as regression analysis.
• Regression analysis is a study of the relationships among variables. It is
used to determine a curve that minimizes the error and provides a good
approximation to the physical problem
Linear regression (simplest form) - using a straight line to fit a set of data
points (x1,y1), (x2, y2), ..., (xn, yn)
where the dependent variable y and the independent variable x have linear
relationship. The equation of the straight line is
y = mx + c + e
c = y-intercept
y*
y
x
Figure 5e Linear regression and error definition
The ‘best’ straight line would be the one that minimize the total
error. Several criteria may be used. However, a more practical
criterion for least-squares approach is to minimize the sum of the
squares of the residuals (Sr), that is
n
minimize Sx = ∑e
i=1
2
i
= ∑ (y
n
i − mxi − ci )
2
i=1
Thus, we need to determine the value of m and c that fulfills the equation. This can
be achieved by differentiating with respect to each coefficient and set the
derivatives equal to zero. The differentiations are not presented here. The resulting
expressions are:
nc + m ∑ xi = ∑ yi
c ∑ xi + m ∑ xi2 = ∑ xi yi
In matrix n ∑ xi c ∑ yi
form=> =
∑ xi ∑ xi2 m
∑ x
i iy
n ∑x y −∑x ∑y
i i i i
m=
These equations are called the normal n∑ x − (∑ x )
2 2
i i
equations. They can be solved using
Cramer’s rule for m and c to obtain
C=
∑y i −m ∑x i
n
Error Quantification in Linear Regression
r=
n ∑ x − (∑ x ) • n ∑ y − (∑ y )
2 2 2 2
i i i i
Interpretation of r,
r=±1 : the straight line is a good fit to the data (i.e all the data lie precisely on the line)
r=0 : the straight line is not a good fit to the data (i.e data scattered randomly and do
not fit a straight line)
For example, when r = 0.8, then 100(0.8)2 = 64 % which may be interpreted as 64% of change in y is due to change in x.
Statistical Analysis
Science of decision making in uncertain situation
∑x i
∑[ ] xi − x 2
0.5
∑ (x − x)
i
2
0.5
22
∑x ∑
2
For small samples: n
i −
xi
∑ (
x i − x )
2
s= 2
: var iance, s =
n (n − 1) n −1
Example : Fit a least square line to following data and find trend value
X 1 2 3 4 5
Y 2 5 3 8 7
There are 5 data points => n = 5. Constructing the table for various terms as
follows.
xi yi x2
i
x y
i i
yi2