LISp-Miner Control Language Reference, version: 27.18.15 of 2 Mar 2022

Codes

ACQuantifierSourceType 
ActionStateType 
ActionType 
AFQuantifierType 
AFQuantifierValueType 
AggregateType 
BoolOpType 
BoolType 
BracketType 
CategorySubType 
CedentType 
CFQuantifierHistogramValueType 
CFQuantifierStepCountType 
CFQuantifierStepSizeType 
CFQuantifierType 
CFQuantifierValueType 
CoefficientType 
CompareType 
DataColumnSubType 
DataCharacterType 
DataTableSubType 
DatePartSubType 
DCQuantifierStepCountType 
DCQuantifierStepSizeType 
DCQuantifierType 
DCQuantifierValueType 
DFQuantifierType 
DFQuantifierValueType 
DKQuantifierType 
DKQuantifierValueType 
FTQuantifierType 
FTQuantifierValueType 
GaceType 
GeoColumnSubType 
GeoDistanceType 
InfinityType 
IntervalValueType 
KLQuantifierKLTableValueType 
KLQuantifierType 
KLQuantifierValueType 
LaqMapType 
LaqPatternGridFlagType 
LaqTaskType 
LiteralType 
MCAlgorithmType 
MCCenterType 
MCDistanceType 
MCLinkageType 
MissingsType 
MutualInfluenceScopeType 
MutualInfluenceType 
MutualInfluenceValidityType 
QuantityBinarizeType 
RMAssignType 
RMCaseType 
RMDistributionType 
RMDiversityType 
RMCharibType 
SDConstructionType 
SDQuantifierSourceType 
TaskSubType 
TestingType 
ValueSubType 


Additional codes

LogVerbosityLevel  How much of the execution logged into history
TargetPlatform  How to run a task
TaskGenerationStatus  Task status codes
HypothesisMutualInfluenceRelationship  Types of relationship between hypothesis and mutual influence domain knowledge

Details

ACQuantifierSourceType

Key Name Note
Before  'State before' frequencies Quantifier test applied to the state-before set data only
After  'State after' frequencies Quantifier test applied to the state-after set data only
DiffAbs  Delta of absolute frequencies values Quantifier test applied to the 4ft-table computed by substracting of corresponding frequencies
DiffRel  Delta of relative frequencies values Quantifier test applied to the 4ft-table computed by substracting of corresponding relative frequencies
ValDiff  Difference of interest-measures Test applied to the difference of interest-measures computed separately from each frequency table
ValDiffAbs  Absolute difference of interest-measures Test applied to the absolute difference of interest-measures computed separately from each frequency table
ValRatio  Ratio of interest-measures Test applied to the ratio of interest-measures computed separately from each frequency table
ValRatioMax  Higher of two ratios of interest-measures Test applied to the higher of two ratios of interest-measures (after/before or before/after)

ActionStateType

Key Name Note
Before  Before State before action
After  After State after action

ActionType

Key Name Note
Stable  Stable Stable with no action
Variable  Variable Variable with action possible

AFQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from source contingency table
Min  Min frequency Minimal frequency from source contingency table
Max  Max frequency Maximal frequency from source contingency table
PImplication  p-Implication a/(a+b) >= p ... at least 100*p [%] of objects satisfying A satisfy also S
AverageDifference  Average Difference Dependence Relative difference in number of objects satisfying S among objects satisfying A to number objects saysfying S in the whole data matrix. Paramater p in <-1;inf), similar to the Lift(X->Y)= P(Y|X)/P(Y). It holds p= P(Y|X)/P(Y)-1.
AboveAverage  Above Average Dependence Among objects satisfying A, there are at least 100*p [%] more objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;inf).
BelowAverage  Below Average Dependence Among objects satisfying A, there are at least 100*p [%] less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
OAD  Outside Average Dependence Among objects satisfying A, there is at least 100*p [%] more or less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
LowerCriticalImplication  Lower Critical Implication -
UpperCriticalImplication  Upper Critical Implication -
DoublePImplication  Double p-Implication -
DoubleLowerCriticalImplication  Double Lower Critical Implication -
DoubleUpperCriticalImplication  Double Upper Critical Implication -
FoundedEquivalence  p-Equivalence -
LowerCriticalEquivalence  Lower Critical Equivalence -
UpperCriticalEquivalence  Upper Critical Equivalence -
SimpleDeviation  Simple Deviation -
Fisher  Fisher quantifier -
ChiSq  Chi-Squared quantifier -
AQ  A-quantifier -
EQ  E-quantifier -
afrequency  a-frequency a-frequency (BASE) from source contingency table

AFQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelMax  Relative [%] to max frequency Relative to maximal value in the current contingency table. Threshold value [%] is multiplied by (the highest frequency)/100

AggregateType

Key Name Note
COUNT  Count(*) Number of rows in group.
SUM  Sum Sum of values in group. Each category is represented by its index.
MIN  Minimum Minimum value in group. Each category is represented by its index.
MAX  Maximum Maximum value in group. Each category is represented by its index.
AVG  Average Average value in group. Each category is represented by its index.

BoolOpType

Key Name Note
Conjunction  Conjunction Logical operation of conjunction
Disjunction  Disjunction Logical operation of disjunction

BoolType

Key Name Note
NotSet  No boolean Not boolean category
True  True Category representing 'TRUE' values
False  False Category representing 'FALSE' values

BracketType

Key Name Note
Sharp  Sharp Include border value
Round  No sharp Don't include border value

CategorySubType

Key Name Note
Enumeration  Enumeration Enumeration of values
Interval  Interval Interval of values
FuzzyInterval  Fuzzy interval Fuzzy intervals

CedentType

Key Name Note
Antecedent  Antecedent Left 4ft-cedent
Succedent  Succedent Right 4ft-cedent
Condition  Condition Condition
FirstSet  First set First set for difference
SecondSet  Second set Second set for difference
AntecedentVar  Variable antecedent Action in antecedent
SuccedentVar  Variable succedent Action in succedent
CFAttribute  Histogram attribute Attribute used for calculation of CF-histogram
KLAttributeRow  Row attribute Row attribute for KL-table
KLAttributeCol  Col attribute Column attribut for KL-table
ETAttribute  ETree Attribute Attribute used in tree
ETClass  Class Attribute Target class attribute
MCAttribute  Vector Attribute Attribute used in clustering
ConditionedBy  MI Conditioned be Condition for a mutual influence truthfullness scope
NotInfluencedBy  MI Not influenced by Condition not influencing a mutual influence truthfullness

CFQuantifierHistogramValueType

Key Name Note
Abs  Absolute number Absolute frequencies. Frequencies in histogram are left as given.
RelCondition  Relative [%] to act condition Relative frequencies to number of rows matching condition. Frequencies are divided by number of rows matching condition and multiplied by 100.
RelCategory  Relative [%] for each category Relative frequencies for each category (number of rows matching condition to number of all). Each frequency is divided by the corresponding frequency in the whole matrix and multiplied by 100.

CFQuantifierStepCountType

Key Name Note
Abs  Absolute number Absolute number of steps. Count value is left as given.
RelRange  Relative [%] to act range Relative to number of categories in the currently selected range- 1. Count value is multiplied by ('number of categories in current range minus 1')/100
RelAll  Relative [%] to all categories Relative to number of all categories minus 1. Count value is multiplied by ('total number of categories minus 1')/100

CFQuantifierStepSizeType

Key Name Note
Abs  Absolute number Change in frequency as an absolute number. Minimal step size is left as given.
RelFrequencyPrev  Relative [%] to previous frequency Relative to the previous frequency. Minimal step size is multiplied by (frequency of the previous category)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Minimal step size is multiplied by (number of rows in the whole matrix)/100
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Minimal step size is multiplied by (number of rows matching condition)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the currently processed histogram. Minimal step size is multiplied by (the highest frequency in the whole histogram)/100

CFQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from range of categories
Min  Min frequency Minimal frequency from range of categories
Max  Max frequency Maximal frequency from range of categories
Avg  Average frequency Average frequency from range of categories
Some  Some frequency At least one frequency from range of categories
VariationRatio  Variation ratio Variation ratio = 1-f(modal)
NominalVariation  Nominal variation (norm) Nominal variation (norm) = suma(f_i*(1-f_i)) * K/(K-1), where f_i is frequency of the i-th category and K is number of categories
DiscreteOrdinaryVariation  Discrete ordinary variation (norm) Discrete Ordinary Variation (norm) = 2*suma(F_i*(1-F_i))* 2/(K-1), where F_i is cumulative relative frequency of the i-th category and K is number of categories
MedianIA  Median-category index (absolute) Index (absolute <1;K>) of the median category
MedianIR  Median-category index (relative) Index (relative <0;1>) of the median category
ArithmeticAverage  Arithmetic average Arithmetic average of cardinal values. Only for cardinal attributes.
GeometricAverge  Geometric average Geometric average of cardinal values. Only for cardinal attributes.
Variance  Variance Variance of cardinal values. Only for cardinal attributes.
StDev  Standard deviation Standard deviation of cardinal values. Only for cardinal attributes.
Skewness  Skewness Skewness of distribution of cardinal values. Only for cardinal attributes.
Asymetry  Asymetry Asymetry coeficient of distribution of cardinal values. Only for cardinal attributes.
StepsUp  Steps-up Number of steps-up in frequency of adjectant categories from given range in histogram
StepsDown  Steps-down Number of steps-down in frequency of adjectant categories from given range in histogram
PattDiffSum  PattDiffSum Sum of absolute values of differences of frequencies in the histogram and a given pattern
PattDiffMin  PattDiffMin The minimal of absolute values of differences of frequencies in the histogram and a given pattern
PattDiffMax  PattDiffMax The maximal of absolute values of differences of frequencies in the histogram and a given pattern
Var  Variation from pattern Total variation from given pattern

CFQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the whole histogram. Threshold value [%] is multiplied by (the highest frequency in the whole histogram)/100

CoefficientType

Key Name Note
Subset  Subsets All the subsets of categories from minimal up to maximal length
OneCategory  One category Only one selected category appears in literal
Sequence  Sequences All the consecutive sequencies of categories from minimal up to maximal length
CyclicalSequence  Cyclical sequences Same as the sequence plus overlapping combinations of first and last categories
Cuts  Cuts All the cuts of categories from minimal up to maximal length
LeftCut  Left cuts All the left cuts of categories from minimal up to maximal length
RightCut  Right cuts All the right cuts of categories from minimal up to maximal length

CompareType

Key Name Note
Equal  Equal Equal value
Less  Less than Less than
LessOrEqual  Less than or equal Less than or equal
Greater  Greater than Greater than
GreaterOrEqual  Greater than or equal Greater than or equal
NotEqual  Not equal Not equal value

DataColumnSubType

Key Name Note
Ordinary  Ordinary DB field Ordinary database column
Derived  Derived Derived column by expression
SQLQuery  SQL-query Column computed as an SQL query
Geo  Geo Column computed as a geographical relation of a point to a set of points or polygons
DatePart  Date-time partial value A single partial value derived from the DateTime column
MCField  Multi-column field One field of a multi-column
Hypothesis  Task-results derived value Derived column based on results of a hypothesis
PCA  Task-results derived value One component of a Principal component analysis

DataCharacterType

Key Name Note
Nominal  Nominal Nominal values
Ordinal  Ordinal Ordinal values with an order defined
Cardinal  Cardinal Cardinal values with a distance defined
NotSet  Not set Unknown character of values

DataTableSubType

Key Name Note
Table  Stored table Data statically stored in database table
View  Dynamic view Dynamically created view from (multiple) database tables

DatePartSubType

Key Name Note
Year  Year Year-value of the date
Month  Month Month-value of the date
Day  Day Day-value from of the date
Hour  Hour Hour-value of the time
Min  Min Minutes-value of the time
Sec  Sec Seconds-value of the time
DayOfWeek  DayOfWeek Day of the week of the date
DayOfYear  DayOfYear Day of the year of the date
WeekOfYear  WeekOfYear Week of the year of the date
Quarter  Quarter Quarter of the year of the date
DayOfRange  DayOfRange Day index within the min-max range of dates in the corresponding column
MonthOfRange  MonthOfRange Month index within the min-max range of dates in the corresponding column
NotSet  Not set Not applicable to this attribute

DCQuantifierStepCountType

Key Name Note
Abs  Absolute number Absolute number of steps. Count value is left as given.
RelRange  Relative [%] to act range Relative to number of categories in the currently selected range- 1. Count value is multiplied by ('number of categories in current range minus 1')/100
RelAll  Relative [%] to all categories Relative to number of all categories minus 1. Count value is multiplied by ('total number of categories minus 1')/100

DCQuantifierStepSizeType

Key Name Note
Abs  Absolute number Change in frequency as an absolute number. Minimal step size is left as given.
RelFrequencyPrev  Relative [%] to previous frequency Relative to the previous frequency. Minimal step size is multiplied by (frequency of the previous category)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Minimal step size is multiplied by (number of rows in the whole matrix)/100
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Minimal step size is multiplied by (number of rows matching condition)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the currently processed histogram. Minimal step size is multiplied by (the highest frequency in the whole histogram)/100

DCQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from range of categories
Min  Min frequency Minimal frequency from range of categories
Max  Max frequency Maximal frequency from range of categories
Avg  Average frequency Average frequency from range of categories
Some  Some frequency At least one frequency from range categories
VariationRatio  Variation ratio Variation ratio = 1-f(modal)
NominalVariation  Nominal variation (norm) Nominal variation (norm) = suma(f(i)*(1-f(i))) * K/(K-1)
DiscreteOrdinaryVariation  Discrete ordinary variation (norm) Discrete Ordinary Variation (norm) = 2*suma(F(i)*(1-F(i))* 2/(K-1)
ArithmeticAverage  Arithmetic average Arithmetic average of cardinal values
GeometricAverage  Geometric average Geometric average of cardinal values
Variance  Variance Variance of cardinal values
StandardDeviation  Standard deviation Standard deviation of cardinal values
Skewness  Skewness Skewness of distribution of cardinal values
Asymetry  Asymetry Asymetry coeficient of distribution of cardinal values
StepsUp  Steps-up Number of steps-up in frequency of adjectant categories in histogram
StepsDown  Steps-down Number of steps-down in frequency of adjectant categories in histogram
Var  Variation from pattern Total variation from given pattern

DCQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal value in the current contingency table. Threshold value [%] is multiplied by (the highest frequency in the whole histogram)/100

DFQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from source contingency table
Min  Min frequency Minimal frequency from source contingency table
Max  Max frequency Maximal frequency from source contingency table
PImplication  p-Implication a/(a+b) >= p ... at least 100*p [%] of objects satisfying A satisfy also S
AverageDifference  Average Difference Dependence Relative difference in number of objects satisfying S among objects satisfying A to number objects saysfying S in the whole data matrix. Paramater p in <-1;inf), similar to the Lift(X->Y)= P(Y|X)/P(Y). It holds p= P(Y|X)/P(Y)-1.
AboveAverage  Above Average Dependence Among objects satisfying A there are at least 100*p [%] more objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;inf).
BelowAverage  Below Average Dependence Among objects satisfying A there are at least 100*p [%] less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
OAD  Outside Average Dependence Among objects satisfying A, there is at least 100*p [%] more or less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
LowerCriticalImplication  Lower Critical Implication The binomical test rejects on the level alpha the null hypothesis P(S|A)<=p in favour of alternative P(S|A)>p
UpperCriticalImplication  Upper Critical Implication The binomical test does not reject on the level alpha the null hypothesis P(S|A)<=p in favour of alternative P(S|A)>p
DoublePImplication  Double p-Implication a/(a+b+c) >= p ... at least 100*p [%] of objects satisfying A or S satisfy both A and S
DoubleLowerCriticalImplication  Double Lower Critical Implication The binomical test rejects on the level alpha the null hypothesis P(AandS|AorS)<=p in favour of alternative P(AandS|AorS)>p
DoubleUpperCriticalImplication  Double Upper Critical Implication The binomical test does not reject on the level alpha the null hypothesis P(AandS|AorS)<=p in favour of alternative P(AandS|AorS)>p
PEquivalence  p-Equivalence (a+d)/n >= p ... at least 100*p [%] objects have the same truth value for A and S
LowerCriticalEquivalence  Lower Critical Equivalence The binomical test rejects on the level alpha the null hypothesis P(A and S have the same truth value)<=p in favour of alternative P(A and S have the same truth value)>p
UpperCriticalEquivalence  Upper Critical Equivalence The binomical test does not reject on the level alpha the null hypothesis P(A and S have the same truth value)<=p in favour of alternative P(A and S have the same truth value)>p
SimpleDeviation  Simple Deviation a*d > exp(sigma)*b*c
Fisher  Fisher quantifier The one-sided Fisher test rejects on the level alpha the null hypothesis of independence of A and S in favour of the alternative of their positive logarithmic interaction
ChiSq  Chi-Squared quantifier The one-sided Fisher test asymptotically rejects on the level alpha the null hypothesis of independence of A and S in favour of the alternative of their positive logarithmic interaction
AQ  A-quantifier -
EQ  E-quantifier -
afrequency  a-frequency a-frequency (BASE) from source contingency table

DFQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal value in the current contingency table. Threshold value [%] is multiplied by (the highest frequency)/100

DKQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from given part of source contingency table
Min  Min frequency Minimal frequency from given part of source contingency table
Max  Max frequency Maximal frequency from given part of source contingency table
Avg  Average frequency Average frequency from given part of source contingency table
Some  Some frequency At least one frequency from given part of source contingency table
CramerV  Cramer's V coefficient Cramer's V (association of two nominal variables) in <0;1> (the farther is value from 0 the more dependant)
Kendall  Kendall's TauB coefficient Value of TauB in <-1;1> (the farther is value from 0 the more dependant)
ChiSq  Chi-square test Chi-square test of similarity (the greater value the more dependant)
ConditionalEntropy  Conditional entropy H(C|R) Conditional entropy of columns given rows (the lower value the more dependant <0;log2(L)>)
MutualInformation  Mutual information MI(R,C) normalized Mutual information between rows and columns (the greater value the more dependant <0;1>)
InformationDependence  Inf. dependence ID(R,C) Information dependence between rows and columns (the greater value the more dependant <0;1>)
AsymetricInformation  Asymetric information coefficient AIC(R,C) Value of asymetric information coefficient Theta (the greater value the more dependant <0;1>)

DKQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the current contingency table. Threshold value [%] is multiplied by (the highest frequency in the whole KL-table)/100

FTQuantifierType

Key Name Note
Support  Support a/(a+b+c+d) >= p ... at least 100*p [%] of objects satisfy both A and S
PImplication  p-Implication a/(a+b) >= p ... at least 100*p [%] of objects satisfying A satisfy also S
AboveAverage  Above Average Dependence Among objects satisfying A, there is at least 100*p [%] more objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;inf).
BelowAverage  Below Average Dependence Among objects satisfying A, there is at least 100*p [%] less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
OAD  Outside Average Dependence Among objects satisfying A, there is at least 100*p [%] more or less objects satisfying S than there are objects satisfying S in the whole data matrix. Parameter p in <0;1>.
AverageDifference  Average Difference Dependence Relative difference in number of objects satisfying S among objects satisfying A to number objects saysfying S in the whole data matrix. Paramater p in <-1;inf), similar to the Lift(X->Y)= P(Y|X)/P(Y). It holds p= P(Y|X)/P(Y)-1.
LowerCriticalImplication  Lower Critical Implication The binomical test rejects on the level alpha the null hypothesis P(S|A)<=p in favour of alternative P(S|A)>p
UpperCriticalImplication  Upper Critical Implication The binomical test does not reject on the level alpha the null hypothesis P(S|A)<=p in favour of alternative P(S|A)>p
DoublePImplication  Double p-Implication a/(a+b+c) >= p ... at least 100*p [%] of objects satisfying A or S satisfy both A and S
DoubleLowerCriticalImplication  Double Lower Critical Implication The binomical test rejects on the level alpha the null hypothesis P(AandS|AorS)<=p in favour of alternative P(AandS|AorS)>p
DoubleUpperCriticalImplication  Double Upper Critical Implication The binomical test does not reject on the level alpha the null hypothesis P(AandS|AorS)<=p in favour of alternative P(AandS|AorS)>p
PEquivalence  p-Equivalence (a+d)/n >= p ... at least 100*p [%] objects have the same truth value for A and S
LowerCriticalEquivalence  Lower Critical Equivalence The binomical test rejects on the level alpha the null hypothesis P(A and S have the same truth value)<=p in favour of alternative P(A and S have the same truth value)>p
UpperCriticalEquivalence  Upper Critical Equivalence The binomical test does not reject on the level alpha the null hypothesis P(A and S have the same truth value)<=p in favour of alternative P(A and S have the same truth value)>p
SimpleDeviation  Simple Deviation a*d > exp(delta)*b*c
Fisher  Fisher quantifier The one-sided Fisher test rejects on the level alpha the null hypothesis of independence of A and S in favour of the alternative of their positive logarithmic interaction
ChiSq  Chi-Square quantifier The Chi-Square test asymptotically rejects on the level alpha the null hypothesis of independence of A and S in favour of the alternative of their positive logarithmic interaction
EQ  E-quantifier -
ParaSeparation  Paraconsistent separation Paraconsistent separation quantifier looking for (almost) separate A and S with truth-criterion of (1+p)*a<=b+c. It is recommend also to set-up a simple-frequency quantifier for (b+c) to assure some minimal sample size.
BASE  BASE a >= BASE ... at least BASE-number of objects for dependency to be statistically relevant
CEIL  Ceiling a <= CEIL ... not more than CEIL-number of objects (i.e. not too 'obvious' dependency)
afrequency  a-frequency a-frequency from contingency table
bfrequency  b-frequency b-frequency from contingency table
cfrequency  c-frequency c-frequency from contingency table
dfrequency  d-frequency d-frequency from contingency table
rfrequency  r-frequency r-frequency (a+b) from contingency table
sfrequency  s-frequency s-frequency (c+d) from contingency table
kfrequency  k-frequency k-frequency (a+c) from contingency table
lfrequency  l-frequency l-frequency (b+d) from contingency table
adfrequency  ad-frequency ad-frequency (a+d) from contingency table
bcfrequency  bc-frequency bc-frequency (b+c) from contingency table
Sum  Sum of values Sum of frequencies (n) from contingency table
Min  Min value Minimal frequency from contingency table
Max  Max value Maximal frequency from contingency table
FreqDiffVal  Frequency difference Frequency of corresponding category in histogram minus number of rows matching both antecedent and succedent
FreqDiffRel  Relative frequencies difference Difference of relative frequency of a corresponding category in histogram and confidence
FreqDiffRelAbs  Absolute value of relative frequencies difference Absolute value of difference of relative frequency of a corresponding category in histogram and confidence
FreqDiffRelRatio  Ratio of relative frequencies Ratio of relative frequency in histogram to relative frequency of succedent in rows matching antecedent (i.e. "inverse Lift")
FreqDiffCat  Relative frequencies of category difference Difference of relative frequency of category in histogram to the frequency of this category in the whole data and relative frequency of rows matching both antecedent and succedent to frequency of category in the whole data
FreqDiffCatAbs  Absolute value of relative frequencies of category difference Absolute value of difference of relative frequency of category in histogram to the frequency of this category in the whole data and relative frequency of rows matching both antecedent and succedent to frequency of category in the whole data
FreqDiffCatRatio  Ratio of relative frequencies of category Ratio of relative frequency of category in histogram to the frequency of this category in the whole data and relative frequency of rows matching both antecedent and succedent to frequency of category in the whole data

FTQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the current contingency table. Threshold value [%] is multiplied by (the highest frequency)/100

GaceType

Key Name Note
Positive  Positive Positive gace
Negative  Negative Negative gace (negation)
Both  Both Both (positive and negative)

GeoColumnSubType

Key Name Note
InPolygonName  In-Polygon name Name of the polygon the coordinates are within
PointName  Closest-Point name Name of the closest point to the coordinates
PointDistance  Closest-Point distance Distance [km] from the coordinates to the the closest point
PointCount  Point-number Number of points within a distance [km] from the coordinates (distance given by a threshold value)
NotSet  Not set Not applicable to this attribute

GeoDistanceType

Key Name Note
Euclidean  Euclidean Euclidean distance of cartesian planar coordinates.
Cylindrical  Cylindrical Euclidean distance on cylindrical projection of geographical coordinates (WGS84) and transformed into kilometers. Applicable for points up to 500 km apart and within latitudes of 70N and 70S.
Haversine  Haversine Haversine formula for computing distances on the sphere for geographical coordinates (WGS84) and transformed into kilometers. The better precison for large distances on the Earth, but slowest.
NotSet  Not set Not applicable to this attribute

InfinityType

Key Name Note
Finite  Finite value Finite value (not infinity)
PlusInfinity  Plus infinity Plus infinity value
MinusInfinity  Minus infinity Minus infinity value

IntervalValueType

Key Name Note
From  From Left border value
To  To Right border value
FuzzyFromMin  Fuzzy from (min) First left fuzzy border value
FuzzyFromMax  Fuzzy from (max) Second left fuzzy border value
FuzzyToMin  Fuzzy to (min) First right fuzzy border value
FuzzyToMax  Fuzzy to (max) Second right fuzzy border value

KLQuantifierKLTableValueType

Key Name Note
Abs  Absolute number Absolute frequencies. Frequencies in histogram are left as given.
RelCondition  Relative [%] to act condition Relative frequencies to number of rows matching condition. Frequencies are divided by number of rows matching condition and multiplied by 100.
RelRow  Relative [%] for each row Relative frequencies for each row. Each frequency is divided by the sum of frequencies in its row and multiplied by 100.
RelCol  Relative [%] for each column Relative frequencies for each column. Each frequency is divided by the sum of frequencies in its column and multiplied by 100.

KLQuantifierType

Key Name Note
Sum  Sum of frequencies Sum of frequencies from given part of contingency table
Min  Min frequency Minimal frequency from given part of contingency table
Max  Max frequency Maximal frequency from given part of contingency table
Avg  Average frequency Average frequency from given part of contingency table
Some  Some frequency At least one frequency from given part of contingency table
CramerV  Cramer's V coefficient Cramer's V (association of two nominal variables) in <0;1> (the farther is value from 0 the more dependant)
Kendall  Kendall's TauB coefficient Kendall's TauB (coefficient of ordinal correlation of two variables) in <-1;1> (the farther is value from 0 the more dependant)
ChiSq  Chi-square test Chi-square test of similarity (the greater value the more dependant)
Variation  Variation from pattern Total variation from pattern
ConditionalEntropy  Conditional entropy H(C|R) Conditional entropy of columns given rows (the lower value the more dependant <0;log2(L)>)
MutualInformation  Mutual information MI(R,C) normalized Mutual information between rows and columns (the greater value the more dependant <0;1>)
InformationDependence  Inf. dependence ID(R,C) Information dependence between rows and columns (the greater value the more dependant <0;1>)
AsymetricInformation  Asymetric information coefficient AIC(R,C) Value of asymetric information coefficient Theta (the greater value the more dependant <0;1>)
PattDiffSum  PattDiffSum Sum of absolute values of differences of frequencies in the contingency table and a given pattern
PattDiffMin  PattDiffMin The minimal of absolute values of differences of frequencies in the contingency table and a given pattern
PattDiffMax  PattDiffMax The maximal of absolute values of differences of frequencies in the contingency table and a given pattern

KLQuantifierValueType

Key Name Note
Abs  Absolute number Absolute number. Threshold value is left as given.
RelCondition  Relative [%] to act condition Relative to number of rows matching condition. Threshold value [%] is multiplied by (number of rows matching condition)/100
RelAll  Relative [%] to all objects Relative to number of all rows in the data matrix. Threshold value [%] is multiplied by (number of rows in the whole matrix)/100
RelFrequencyMax  Relative [%] to max frequency Relative to maximal frequency in the whole contingency table. Threshold value [%] is multiplied by (the highest frequency in the whole KL-table)/100

LaqMapType

Key Name Note
Antecedent  Antecedent Mapping attributes to antecedent
Succedent  Succedent Mapping attributes to succedent
Condition  Condition Mapping attributes to condition
FirstSet  FirstSet Mapping attributes to the firstset
SecondSet  SecondSet Mapping attributes to the secondset
Quantifier  Quantifier Mapping of quantifier
Filter  Filter Mapping of filtering

LaqPatternGridFlagType

Key Name Note
NotSet  Not set Influence not set yet
DontCare  Don't care Not interested in this combination
ToBeMined  To be mined Waiting for to be mined
Processed  Being processed Beeing processed
Solved  Solved Results obtained

LaqTaskType

Key Name Note
FTMiner  4ftMiner 4ftMiner Task
CFMiner  CF-Miner CF-Miner Task
KLMiner  KL-Miner KL-Miner Task
SD4ftMiner  SD4ft-Miner SD4ft-Miner Task
SDCFMiner  SDCF-Miner SDCF-Miner Task
SDKLMiner  SDKL-Miner SDKL-Miner Task

LiteralType

Key Name Note
Basic  Basic Literal must be included in every cedent
Remaining  Remaining Literal is obligatory in cedent

MCAlgorithmType

Key Name Note
KMeans  k-Means / k-Mode Flat level k-Means or k-Mode clustering
RBKMeans  Repeated bisection k-Means / k-Mode Top-down recursive bi-section clustering producing a tree of clusters
HAC  Hierarchical Agglomerative Clustering Bottom-up hierarchical agglomerative clustering using distance matrix

MCCenterType

Key Name Note
CatMean  Mean Mean of category-index values
CatModus  Modus The most frequent category

MCDistanceType

Key Name Note
Euclidean  Euclidean distance Euclidean measure of distance between two objects
Cosinus  Cosinus similarity Cosinus measure of similarity between two objects
SimpleMatch  Simple match Simple match of categories (Sokal and Michener)
Eskin  Eskin measure Esking similarity for nominal variables (Eskin et al.)
VariableEntropy  Variable Entropy Variable Entropy similarity measure of nominal variables (Sulc)
VariableMutability  Variable Mutability Variable Mutability similarity measure of nominal variables (Sulc)

MCLinkageType

Key Name Note
Simple  Simple Simple linkage of clusters (the minimal distance)
Complete  Complete Complete linkage of clusters (the maximal distance)
Average  Average Average linkage of clusters (the average distance)

MissingsType

Key Name Note
Deleting  Delete Not including missing values
Pesimistic  Pesimistic fill up Pesimistic fill up of missing values
Optimistic  Optimistic fill up Pesimistic fill up of missing values
Ignore  Ignore X-categories Ignoring X-categories

MutualInfluenceScopeType

Key Name Note
NotSet  Not set Scope not set yet
BackgroundKnowledge  Background knowledge Dependency follows from the domain background knowledge
DataSpecific  Data specific Dependency follows from data anomalies
Unknown  Unknown Scope is unknown

MutualInfluenceType

Key Name Note
NotSet  Not set Influence not set yet
SomeInfluence  Some influence There is some influence but not examined in detail
PositiveInfluence  Positive influence If the row attribute increases then the column attribute increases too
NegativeInfluence  Negative influence If the row attribute increases then the column attribute decreases
PositiveFrequency  Positive frequency If the row attribute increases then the relative frequency of objects satisfying column attribute increases
NegativeFrequency  Negative frequency If the row attribute increases then the relative frequency of objects satisfying column attribute decreases
PositiveBoolean  Positive boolean If truthfulness of the row attribute increases then relative frequency of true values of column attribute increases too
NegativeBoolean  Negative boolean If truthfulness of the row attribute increases then relative frequency of true values of column attribute decreases
Functional  Functional There is a strong function-like dependency
None  None No influence at all
DontCare  Don't care about influence There maybe some influence but we don't care
Unknown  Unknown There could be an influence, no details are known

MutualInfluenceValidityType

Key Name Note
NotSet  Not set Validity not set yet
Proven  Proven Dependency proven
Rejected  Rejected Dependency was rejected
Unknown  Unknown Validity not know

QuantityBinarizeType

Key Name Note
None  None Not set
EqualNotEqual  Equal versus Not-equal One category against all others. As many dichotimized attributes with categories '=' and '<>' as there is categories in the original attribute. Suitable for nominal attributes.
UpToAbove  Up-to versus Above Disjunction of all categories up to a certain category against all other categories. As many dichotomized attributes with categories '<=' and '>' as there is categories in the original attribute minus one. Suitable for ordinal and cardinal attributes.

RMAssignType

Key Name Note
Random  Random Column values generated randomly
Init  Initialized by formula Column values initially computed by a mathematical formula
Update  Updated by formula Column values continuously updated by a mathematical formula
Copy  Copied from the underlying data Column values initialized by copying values from the underlying data

RMCaseType

Key Name Note
Evolution  Evolution Generating artificial data using evolutionary algorithm
Randomizer  Randomizer Randomization of previously generated data using evolution algorithm

RMDistributionType

Key Name Note
Uniform  Uniform Uniform distribution of values across possible range
Gaussian  Gaussian A bell-shaped distribution with defined mean value and standard deviation
Manually  Manually user-specified Manually specified frequencies (for enumerations only)

RMDiversityType

Key Name Note
Random  Random Randomly generated values for each individual
Permutation  Permutation Permutation of a one-time pregenerated random values

RMCharibType

Key Name Note
Enumeration  Enumeration Column could contain values only from a pre-defined list
Range  Range of values Column with any value within a given range
RangeEnum  Range with predefined values Column with any value within a given range with some predefined values

SDConstructionType

Key Name Note
Simple  FirstSet versus SecondSet Comparison of the first set versus the second set
JoinedSet  FirstSet versus FirstSet and SecondSet The second set is treated as a subset specification to the first set
InverseSet  FirstSet versus not FirstSet The second set is inverse of the first set

SDQuantifierSourceType

Key Name Note
FirstSet  First set frequencies Quantifier test applied to the first set data only
SecondSet  Second set frequencies Quantifier test applied to the second set data only
DiffAbs  Delta of absolute frequencies values Quantifier test applied to the frequency table computed by substracting of corresponding frequencies
DiffRel  Delta of relative frequencies values Quantifier test applied to the frequency table computed by substracting of corresponding relative frequencies
ValDiff  Difference of interest-measures Test applied to the difference of quantifier values computed separately from each frequency table
ValDiffAbs  Absolute difference of interest-measures Test applied to the absolute difference of interest-measures computed separately from each frequency table
ValRation  Ratio of interest-measures Test applied to the ratio of interest-measures computed separately from each frequency table
ValRatioMax  Higher of ratios of interest-measures Test applied to the higher of two ratios of interest-measures (im1/im2 or im2/im1)

TaskSubType

Key Name Note
FTMiner  4ft-Miner 4ft-Miner task for 4ft-association rules with rich syntax
CFMiner  CF-Miner CF-Miner task for conditonal frequencies of a single multi-categorical attribute
KLMiner  KL-Miner KL-Miner task for conditional frequencies of two multi-categorical attributes
SD4ftMiner  SD4ft-Miner SD4ft-Miner task for set difference in terms of two 4ft-association rules
SDCFMiner  SDCF-Miner SDCF-Miner task for set difference in terms of two frequencies of a single attribute
SDKLMiner  SDKL-Miner SDKL-Miner task for set difference in terms of two frequencies of two multi-categorical attributes
Ac4ftMiner  Ac4ft-Miner Ac4ft-Miner task for 4ft-action rules
ETreeMiner  ETree-Miner ETree-Miner task for decision and exploration trees
MClusterMiner  MCluster-Miner MCluster-Miner task for clustering analysis
KEx  Knowledge Explorer Knowledge Explorer (KEx) task for classifications using machine learning

TestingType

Key Name Note
TrainingSet  Use training set Use same records as for training
CrossValidation  Cross-validation Cross-validating each of n-folds
RandomSplit  Random split Random split of available data in given ration

ValueSubType

Key Name Note
Integer  Integer number Integer number between -2E09 and 2E09
Float  Decimal number Floating-point number with double precision
String  Text String of characters
Boolean  Boolean Boolean value of TRUE/FALSE
DateTime  Date/Time Value with date, time or both


Additional codes

LogVerbosityLevel

Key Name Note
Error Error Errors only
Warning Warning Include warnings
Info Info Include calls to lm.logInfo
Normal Normal Include all calls to lm.log
Fine Fine Log calls of main functions
Finer Finer Log calls of all functions with named paremeters
Finest Finest for debugging purposes

TargetPlatform

Key Name Note
TaskPooler TaskPooler Add task to TaskPooler queue (no parallelization)
ProcPooler ProcPooler Add task to ProcPooler queue (parallel procesing on multiple cores)
GridPooler GridPooler Add task to GridPooler queue (distributed parallel processing)

TaskGenerationStatus

Key Name Note
None Not generated yet For newly created tasks or after a change has been made to task description
Waiting Waiting Waiting in the queue of TaskPooler, ProcPooler or GridPooler
Running Running Running locally or on the grid
Solved Solved Successfully finished
Interrupted Interrupted Finished because of the HypothesesCountMax was reached or interrupted by user
Failed Failed An error occured which caused the task run to stop

HypothesisMutualInfluenceRelationship

Key Name Note
Unknown Unknown Relationship between the hypothesis and given mutual influence could not be described
Unrelated Unrelated The hypothesis and given mutual influence are not related
DerivedFrom DerivedFrom Hypothesis is derived from the mutual influence (supports it)
Extending Extending Hypothesis extends the mutual influence
InConflictWith InConflictWith Hypothesis is in conflict with the mutual influence (is an exception to)