4385132063

Build # 4385132063

Build Type

push

github

Committed by GitHub

Commit Message

Merge pull request #316 from jecisc/divers-cleanings

Run Details

2977 of 2977 new or added lines in 214 files covered. (100.0%)

19725 of 24212 relevant lines covered (81.47%)

2.44 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

70.27

/src/Math-Numerical/PMMaximumLikelihoodHistogramFit.class.st

Class {
        #name : #PMMaximumLikelihoodHistogramFit,
        #superclass : #PMLeastSquareFit,
        #instVars : [
                'count',
                'countVariance'
        ],
        #category : #'Math-Numerical'
}

{ #category : #operation }
PMMaximumLikelihoodHistogramFit >> accumulate: aWeightedPoint [
                "Private"
        | f g temp inverseProbability|
        f := result valueAndGradient: aWeightedPoint xValue.
        g := f last copyFrom: 1 to: ( f last size - 1).
        f := f first.
        f = 0 ifTrue: [ ^nil].
        inverseProbability := 1 / f.
        temp := aWeightedPoint yValue * inverseProbability.
        constants accumulate: g * temp.
        temp := temp * inverseProbability.
        1 to: g size do:
                [ :k |
                  ( equations at: k) accumulate: g * ( ( g at: k) * temp).
                ]
]

{ #category : #operation }
PMMaximumLikelihoodHistogramFit >> computeChanges [
                "Private"
        ^super computeChanges copyWith: 0
]

{ #category : #operation }
PMMaximumLikelihoodHistogramFit >> computeNormalization [
        "Private"

        | numerator denominator temp |
        numerator := 0.
        denominator := 0.
        dataHolder pointsAndErrorsDo:
                        [:each |
                        temp := result value: each xValue.
                        temp = 0
                                ifFalse:
                                        [numerator := numerator + (each yValue squared / temp).
                                        denominator := denominator + temp]].
        count := ( numerator / denominator) sqrt.
        countVariance := numerator / ( 4 * count)
]

{ #category : #operation }
PMMaximumLikelihoodHistogramFit >> finalizeIterations [
        "Compute the normalization factor."

        self computeNormalization.
        result setCount: count.
        super finalizeIterations
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> fitType [
        ^ 'Maximum likelihood fit'
]

{ #category : #operation }
PMMaximumLikelihoodHistogramFit >> initializeIterations [
        result setCount: 1.
        count := dataHolder totalCount.
        super initializeIterations
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> normalization [
        ^ count
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> normalizationError [
        ^countVariance sqrt
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> numberOfFreeParameters [
                "Private"
        ^super numberOfParameters
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> numberOfParameters [
                "Private"
        ^super numberOfParameters - 1
]

{ #category : #information }
PMMaximumLikelihoodHistogramFit >> valueAndError: aNumber [

        | valueGradient gradient gVar |
        valueGradient := result valueAndGradient: aNumber.
        gradient := valueGradient last copyFrom: 1 to: valueGradient last size - 1.
        gVar := gradient * (self errorMatrix * gradient) / count.
        ^Array with: valueGradient first
                with: ((valueGradient first / count) squared * countVariance + gVar) sqrt
]

1	Class {
2	#name : #PMMaximumLikelihoodHistogramFit,
3	#superclass : #PMLeastSquareFit,
4	#instVars : [
5	'count',
6	'countVariance'
7	],
8	#category : #'Math-Numerical'
9	}
10
11	{ #category : #operation }
12	PMMaximumLikelihoodHistogramFit >> accumulate: aWeightedPoint [	3✔
13	"Private"	3✔
14	\| f g temp inverseProbability\|	3✔
15	f := result valueAndGradient: aWeightedPoint xValue.	3✔
16	g := f last copyFrom: 1 to: ( f last size - 1).	3✔
17	f := f first.	3✔
18	f = 0 ifTrue: [ ^nil].	3✔
19	inverseProbability := 1 / f.	3✔
20	temp := aWeightedPoint yValue * inverseProbability.	3✔
21	constants accumulate: g * temp.	3✔
22	temp := temp * inverseProbability.	3✔
23	1 to: g size do:	3✔
24	[ :k \|	3✔
25	( equations at: k) accumulate: g * ( ( g at: k) * temp).	3✔
26	]	3✔
27	]	3✔
28
29	{ #category : #operation }
30	PMMaximumLikelihoodHistogramFit >> computeChanges [	3✔
31	"Private"	3✔
32	^super computeChanges copyWith: 0	3✔
33	]	3✔
34
35	{ #category : #operation }
36	PMMaximumLikelihoodHistogramFit >> computeNormalization [	3✔
37	"Private"	3✔
38		3✔
39	\| numerator denominator temp \|	3✔
40	numerator := 0.	3✔
41	denominator := 0.	3✔
42	dataHolder pointsAndErrorsDo:	3✔
43	[:each \|	3✔
44	temp := result value: each xValue.	3✔
45	temp = 0	3✔
46	ifFalse:	3✔
47	[numerator := numerator + (each yValue squared / temp).	3✔
48	denominator := denominator + temp]].	3✔
49	count := ( numerator / denominator) sqrt.	3✔
50	countVariance := numerator / ( 4 * count)	3✔
51	]	3✔
52
53	{ #category : #operation }
54	PMMaximumLikelihoodHistogramFit >> finalizeIterations [	3✔
55	"Compute the normalization factor."	3✔
56		3✔
57	self computeNormalization.	3✔
58	result setCount: count.	3✔
59	super finalizeIterations	3✔
60	]	3✔
61
62	{ #category : #information }
63	PMMaximumLikelihoodHistogramFit >> fitType [	×
64	^ 'Maximum likelihood fit'	×
65	]	×
66
67	{ #category : #operation }
68	PMMaximumLikelihoodHistogramFit >> initializeIterations [	3✔
69	result setCount: 1.	3✔
70	count := dataHolder totalCount.	3✔
71	super initializeIterations	3✔
72	]	3✔
73
74	{ #category : #information }
75	PMMaximumLikelihoodHistogramFit >> normalization [	×
76	^ count	×
77	]	×
78
79	{ #category : #information }
80	PMMaximumLikelihoodHistogramFit >> normalizationError [	×
81	^countVariance sqrt	×
82	]	×
83
84	{ #category : #information }
85	PMMaximumLikelihoodHistogramFit >> numberOfFreeParameters [	×
86	"Private"	×
87	^super numberOfParameters	×
88	]	×
89
90	{ #category : #information }
91	PMMaximumLikelihoodHistogramFit >> numberOfParameters [	3✔
92	"Private"	3✔
93	^super numberOfParameters - 1	3✔
94	]	3✔
95
96	{ #category : #information }
97	PMMaximumLikelihoodHistogramFit >> valueAndError: aNumber [	×
98		×
99	\| valueGradient gradient gVar \|	×
100	valueGradient := result valueAndGradient: aNumber.	×
101	gradient := valueGradient last copyFrom: 1 to: valueGradient last size - 1.	×
102	gVar := gradient * (self errorMatrix * gradient) / count.	×
103	^Array with: valueGradient first	×
104	with: ((valueGradient first / count) squared * countVariance + gVar) sqrt	×
105	]	×

PolyMathOrg / PolyMath / 4385132063

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous