ImageJ

curve fitting

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

3 messages Options

Charlie MAO

curve fitting

Hello all,
I am trying to write a plugin in Image J that will use "curve
fitting" routine. But I got wrong result. Then I used following data and
opened in "Curve Fitting" window, select "Straight line" option. I obtained
the following result

Number of iterations: 60
Maximum number of iterations: 2000
Sum of residuals squared: 0.0
Standard deviation: 0.0
Goodness of fit: 0.0
Parameters:
a = 0.0
b = 1.0

448.13,199
446.89,198
446.71,197
447.36,196
446.17,195
448.1,194
446.46,193
447.7,192
446.91,191
448,190
447.09,189
446.45,188
447.92,187
446.92,186
447.58,185
447.09,184
451.14,183
451,182
451.5,181
452,180
451.5,179
451.86,178
452,177
451.5,176
454.27,175
455.17,174
455.09,173
456.08,172
455.5,171
455.38,170
455,169
456,168
455.91,167
455.36,166
455.9,165
455,164
456.33,163
455.73,162
457.1,161
456.8,160
462.71,159
463.44,158
462.67,157
462.71,156
464.08,155
463,154
463.38,153
463.08,152
462.91,151
464,150
463.09,149
462.45,148
463.92,147
462.92,146
463.58,145
463.09,144
467.14,143
467,142
467.5,141
468,140
467.5,139
467.86,138
468,137
467.5,136
472.08,135
470.79,134
472,133
471.45,132
472.42,131
472.18,130
471.07,129
471.75,128
471.85,127
471,126
471.64,125
471.75,124
472.17,123
472.36,122
472.5,121
473.6,120
479.5,119
478,118
478.75,117
480.08,116
478.73,115
479.83,114
479,113
479.75,112
479.18,111
477.78,110
481.33,109
478.91,108
479.58,107
479.33,106
479.2,105
480.14,104
482.5,103
483,102
484,101
483.14,100
483.5,99
483.5,98
483.86,97
483.71,96
488.08,95
486.79,94
488,93
487.45,92
488.42,91
488.18,90
487.07,89
487.75,88
487.85,87
487,86
487.64,85
487.75,84
488.17,83
488.36,82
488.5,81
489.6,80
495.5,79
494,78
494.75,77
496.08,76
494.73,75
495.83,74
495,73
495.75,72
495.25,71
495.93,70
494.82,69
494.58,68
495.55,67
495,66
496.21,65
494.92,64
499.29,63
499.14,62
499.5,61
499.5,60
499.86,59
499,58
500,57
500.5,56
502.86,55
503.8,54
503.67,53
503.42,52
504.09,51
501.67,50
505.22,49
503.82,48
503.25,47
504,46
503.17,45
504.27,44
502.92,43
504.25,42
505,41
503.5,40
509.4,39
510.5,38
510.64,37
510.83,36
511.25,35
511.36,34
512,33
511.15,32
511.25,31
511.93,30
510.82,29
510.58,28
511.55,27
511,26
512.21,25
510.92,24
515.29,23
515.14,22
515.5,21
515.5,20
515.86,19
515,18
516,17
516.5,16
518.23,15
519.8,14
518.83,13
518.92,12
519.1,11
520.2,10
518.89,9
520,8

Any help is appreciated. Thanks

Charlie M MAO, PhD
Senior Research Physicist
RLA
Australia

Wayne Rasband

Re: curve fitting

The Analyze>Tools>Curve Fitting tool expects the data to be space or
tab delimited, not comma delimited.

-wayne

> I am trying to write a plugin in Image J that will use "curve
> fitting" routine. But I got wrong result. Then I used following data
> and
> opened in "Curve Fitting" window, select "Straight line" option. I
> obtained
> the following result
>
> Number of iterations: 60
> Maximum number of iterations: 2000
> Sum of residuals squared: 0.0
> Standard deviation: 0.0
> Goodness of fit: 0.0
> Parameters:
> a = 0.0
> b = 1.0
>
> 448.13,199
> 446.89,198
> 446.71,197
> 447.36,196
> 446.17,195
> 448.1,194
> 446.46,193
> 447.7,192
> 446.91,191
> 448,190
> 447.09,189
> 446.45,188
> 447.92,187
> 446.92,186
> 447.58,185
> 447.09,184
> 451.14,183
> 451,182
> 451.5,181
> 452,180
> 451.5,179
> 451.86,178
> 452,177
> 451.5,176
> 454.27,175
> 455.17,174
> 455.09,173
> 456.08,172
> 455.5,171
> 455.38,170
> 455,169
> 456,168
> 455.91,167
> 455.36,166
> 455.9,165
> 455,164
> 456.33,163
> 455.73,162
> 457.1,161
> 456.8,160
> 462.71,159
> 463.44,158
> 462.67,157
> 462.71,156
> 464.08,155
> 463,154
> 463.38,153
> 463.08,152
> 462.91,151
> 464,150
> 463.09,149
> 462.45,148
> 463.92,147
> 462.92,146
> 463.58,145
> 463.09,144
> 467.14,143
> 467,142
> 467.5,141
> 468,140
> 467.5,139
> 467.86,138
> 468,137
> 467.5,136
> 472.08,135
> 470.79,134
> 472,133
> 471.45,132
> 472.42,131
> 472.18,130
> 471.07,129
> 471.75,128
> 471.85,127
> 471,126
> 471.64,125
> 471.75,124
> 472.17,123
> 472.36,122
> 472.5,121
> 473.6,120
> 479.5,119
> 478,118
> 478.75,117
> 480.08,116
> 478.73,115
> 479.83,114
> 479,113
> 479.75,112
> 479.18,111
> 477.78,110
> 481.33,109
> 478.91,108
> 479.58,107
> 479.33,106
> 479.2,105
> 480.14,104
> 482.5,103
> 483,102
> 484,101
> 483.14,100
> 483.5,99
> 483.5,98
> 483.86,97
> 483.71,96
> 488.08,95
> 486.79,94
> 488,93
> 487.45,92
> 488.42,91
> 488.18,90
> 487.07,89
> 487.75,88
> 487.85,87
> 487,86
> 487.64,85
> 487.75,84
> 488.17,83
> 488.36,82
> 488.5,81
> 489.6,80
> 495.5,79
> 494,78
> 494.75,77
> 496.08,76
> 494.73,75
> 495.83,74
> 495,73
> 495.75,72
> 495.25,71
> 495.93,70
> 494.82,69
> 494.58,68
> 495.55,67
> 495,66
> 496.21,65
> 494.92,64
> 499.29,63
> 499.14,62
> 499.5,61
> 499.5,60
> 499.86,59
> 499,58
> 500,57
> 500.5,56
> 502.86,55
> 503.8,54
> 503.67,53
> 503.42,52
> 504.09,51
> 501.67,50
> 505.22,49
> 503.82,48
> 503.25,47
> 504,46
> 503.17,45
> 504.27,44
> 502.92,43
> 504.25,42
> 505,41
> 503.5,40
> 509.4,39
> 510.5,38
> 510.64,37
> 510.83,36
> 511.25,35
> 511.36,34
> 512,33
> 511.15,32
> 511.25,31
> 511.93,30
> 510.82,29
> 510.58,28
> 511.55,27
> 511,26
> 512.21,25
> 510.92,24
> 515.29,23
> 515.14,22
> 515.5,21
> 515.5,20
> 515.86,19
> 515,18
> 516,17
> 516.5,16
> 518.23,15
> 519.8,14
> 518.83,13
> 518.92,12
> 519.1,11
> 520.2,10
> 518.89,9
> 520,8
>
> Any help is appreciated. Thanks
>
> Charlie M MAO, PhD
> Senior Research Physicist
> RLA
> Australia

Charlie MAO

Re: curve fitting

Thank you Wayne.

I have another question regarding "Goodness of fit". This data set comes
from real problem I like to have a linear fit to find out the slope of the
line in my plugin. I use "Goodness of fit" to validate it. This is the
result:

Number of iterations: 66
Maximum number of iterations: 2000
Sum of residuals squared: 34642.7235318219
Standard deviation: 107.4596413106209
Goodness of fit: 0.9999952402375941
Parameters:
a = 448.12729522146867
b = -9.858691629206724

By visual observation it is not a valid curve fit, but it is a good fit by
the "Goodness of fit" number. Can you tell me what is my problem? Or can I
use "Goodness of fit" to validate the curve fit?

Charlie

N.B. Data set is below:
39 99
39.22 98
38.5 97
39.33 96
38.3 95
37.78 94
37.78 93
38 92
39.6 91
39.22 90
38.63 89
39 88
38.7 87
38.7 86
39.5 85
39.5 84
40.22 83
39 82
39.27 81
38.9 80
40 79
41 78
40.36 77
39.82 76
38.13 75
38.8 74
39.42 73
38.73 72
38.88 71
38.63 70
38.38 69
38.78 68
39.18 67
39.2 66
39.22 65
39.55 64
40.56 63
39.33 62
38.75 61
38.18 60
39.63 59
39.3 58
38.78 57
40.27 56
40 55
39.3 54
39.3 53
39.45 52
39.7 51
38.67 50
39.1 49
39.88 48
40.6 47
39.11 46
40.22 45
38.7 44
39.42 43
39.8 42
39.7 41
38.91 40
40 39
39.33 38
40 37
38.3 36
39.69 35
39.67 34
40.1 33
40.36 32
39 31
38.25 30
38.25 29
39.1 28
40.67 27
40.67 26
39.8 25
39.5 24
39 23
41.5 22
43.5 21
44 20

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]]On Behalf Of Wayne
Rasband
Sent: Tuesday, 15 November 2005 9:45 AM
To: [hidden email]
Subject: Re: curve fitting

The Analyze>Tools>Curve Fitting tool expects the data to be space or
tab delimited, not comma delimited.

-wayne

------------

IMPORTANT: This e-mail and any attachments include privileged and/or
confidential information. If you are not an intended recipient, you must not
copy, distribute or take any action in reliance on it. If you have received
this e-mail in error please notify us immediately by a collect telephone call
to (08) 8352 1322, international + 61 8 8352 1322 or by e-mail at
[hidden email], and delete this e-mail from your system. RLA accepts no
liability for any damage caused by any virus transmitted by this e-mail.