Step-by-step derivation
S001 · implies
Before
source roles = [qf-variance-shift, [F, S, c, denominator, operation, values, weights]]
After
derived = [variance, shift_variance, null]
Conditions: none
Use THIS bank’s actual fields,operation/domain and types.BASE retains permissible extra metadata.Do not import LDL SAFE/set limits or a different bank’s deduplication rules.
Step identifiers and dependencies
Rule: R70_SOURCE_bank_specific_interface · Uses: G001
S002 · calculation
Before
Variance ORIGINAL nonnegative RAW weights sum := sum(1, 8, 5, 0)
After
Variance ORIGINAL nonnegative RAW weights sum = 14
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S001
S003 · implies
Before
source roles = [[0, 1, 2, 3], powerset(S), [-4, -4, 4, 1005], [1, 8, 5, 0], 14, null, -6]
After
derived = [Yes]
Conditions: none
Type ALL RAW values/weights EVEN at zero weight.S is nonempty/distinct,F is powerset(S),weights are nonnegative and sum to q>0.Keep values/weights aligned to ORIGINAL S slots;do not sort independently.Type RAW c/k as well.
Step identifiers and dependencies
Rule: Variance_WHOLE_normalized_model_ORIGINAL_slot_alignment · Uses: G001, S002
S004 · calculation
Before
Variance translated ORIGINAL value i=0 := sum(-4, -6)
After
Variance translated ORIGINAL value i=0 = -10
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S003
S005 · calculation
Before
Variance weight times transformed value i=0 := mul(1, -10)
After
Variance weight times transformed value i=0 = -10
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S004
S006 · implies
Before
source roles = [0, 0, -4, 1, null, -6]
After
derived = [-10, 1/14]
Conditions: none
For EACH ORIGINAL slot,set Y=X,X+c or kX while probability remains w/q.Derived values may exceed SAFE and are not recapped as RAW.Zero-weight slots remain in the source but contribute nothing.
Step identifiers and dependencies
Rule: Variance_EACH_slot_transformation_preserves_probability · Uses: G001, S005
S007 · calculation
Before
Variance translated ORIGINAL value i=1 := sum(-4, -6)
After
Variance translated ORIGINAL value i=1 = -10
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S006
S008 · calculation
Before
Variance weight times transformed value i=1 := mul(8, -10)
After
Variance weight times transformed value i=1 = -80
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S007
S009 · implies
Before
source roles = [1, 1, -4, 8, null, -6]
After
derived = [-10, 4/7]
Conditions: none
For EACH ORIGINAL slot,set Y=X,X+c or kX while probability remains w/q.Derived values may exceed SAFE and are not recapped as RAW.Zero-weight slots remain in the source but contribute nothing.
Step identifiers and dependencies
Rule: Variance_EACH_slot_transformation_preserves_probability · Uses: G001, S008
S010 · calculation
Before
Variance translated ORIGINAL value i=2 := sum(4, -6)
After
Variance translated ORIGINAL value i=2 = -2
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S009
S011 · calculation
Before
Variance weight times transformed value i=2 := mul(5, -2)
After
Variance weight times transformed value i=2 = -10
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S010
S012 · implies
Before
source roles = [2, 2, 4, 5, null, -6]
After
derived = [-2, 5/14]
Conditions: none
For EACH ORIGINAL slot,set Y=X,X+c or kX while probability remains w/q.Derived values may exceed SAFE and are not recapped as RAW.Zero-weight slots remain in the source but contribute nothing.
Step identifiers and dependencies
Rule: Variance_EACH_slot_transformation_preserves_probability · Uses: G001, S011
S013 · calculation
Before
Variance translated ORIGINAL value i=3 := sum(1005, -6)
After
Variance translated ORIGINAL value i=3 = 999
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S012
S014 · calculation
Before
Variance weight times transformed value i=3 := mul(0, 999)
After
Variance weight times transformed value i=3 = 0
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S013
S015 · implies
Before
source roles = [3, 3, 1005, 0, null, -6]
After
derived = [999, 0]
Conditions: none
For EACH ORIGINAL slot,set Y=X,X+c or kX while probability remains w/q.Derived values may exceed SAFE and are not recapped as RAW.Zero-weight slots remain in the source but contribute nothing.
Step identifiers and dependencies
Rule: Variance_EACH_slot_transformation_preserves_probability · Uses: G001, S014
S016 · calculation
Before
Variance mean numerator sum wY := sum(-10, -80, -10, 0)
After
Variance mean numerator sum wY = -100
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S015
S017 · calculation
Before
Variance exact expectation M over q := div(-100, 14)
After
Variance exact expectation M over q = -50/7
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S016
S018 · calculation
Before
Variance i=0 Y times q := mul(-10, 14)
After
Variance i=0 Y times q = -140
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S017
S019 · calculation
Before
Variance i=0 centered integer numerator := sub(-140, -100)
After
Variance i=0 centered integer numerator = -40
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S018
S020 · calculation
Before
Variance i=0 centered numerator squared := mul(-40, -40)
After
Variance i=0 centered numerator squared = 1600
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S019
S021 · calculation
Before
Variance i=0 weight times squared numerator := mul(1, 1600)
After
Variance i=0 weight times squared numerator = 1600
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S020
S022 · calculation
Before
Variance i=1 Y times q := mul(-10, 14)
After
Variance i=1 Y times q = -140
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S021
S023 · calculation
Before
Variance i=1 centered integer numerator := sub(-140, -100)
After
Variance i=1 centered integer numerator = -40
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S022
S024 · calculation
Before
Variance i=1 centered numerator squared := mul(-40, -40)
After
Variance i=1 centered numerator squared = 1600
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S023
S025 · calculation
Before
Variance i=1 weight times squared numerator := mul(8, 1600)
After
Variance i=1 weight times squared numerator = 12800
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S024
S026 · calculation
Before
Variance i=2 Y times q := mul(-2, 14)
After
Variance i=2 Y times q = -28
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S025
S027 · calculation
Before
Variance i=2 centered integer numerator := sub(-28, -100)
After
Variance i=2 centered integer numerator = 72
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S026
S028 · calculation
Before
Variance i=2 centered numerator squared := mul(72, 72)
After
Variance i=2 centered numerator squared = 5184
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S027
S029 · calculation
Before
Variance i=2 weight times squared numerator := mul(5, 5184)
After
Variance i=2 weight times squared numerator = 25920
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S028
S030 · calculation
Before
Variance i=3 Y times q := mul(999, 14)
After
Variance i=3 Y times q = 13986
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S029
S031 · calculation
Before
Variance i=3 centered integer numerator := sub(13986, -100)
After
Variance i=3 centered integer numerator = 14086
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S030
S032 · calculation
Before
Variance i=3 centered numerator squared := mul(14086, 14086)
After
Variance i=3 centered numerator squared = 198415396
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S031
S033 · calculation
Before
Variance i=3 weight times squared numerator := mul(0, 198415396)
After
Variance i=3 weight times squared numerator = 0
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S032
S034 · calculation
Before
Variance sum ALL weighted squared numerators := sum(1600, 12800, 25920, 0)
After
Variance sum ALL weighted squared numerators = 40320
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S033
S035 · calculation
Before
Variance denominator q squared := mul(14, 14)
After
Variance denominator q squared = 196
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S034
S036 · calculation
Before
Variance denominator q cubed binary second factor := mul(196, 14)
After
Variance denominator q cubed binary second factor = 2744
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S035
S037 · calculation
Before
Variance probability population centered sum over q cubed := div(40320, 2744)
After
Variance probability population centered sum over q cubed = 720/49
Conditions: none
This step displays the exact original operands and operation.Mul/sub/div have EXACTLY TWO inputs.Keep rational values exact and derived BigInts unbounded;do not force an answer by approximation.
Step identifiers and dependencies
Rule: exact_rational_arithmetic · Uses: G001, S036
S038 · implies
Before
source roles = [[-10, -10, -2, 999], [1, 8, 5, 0], 14, -50/7, 40320, 2744]
After
derived = [720/49, [-10, -2], No]
Conditions: none
E[Y]=M/q and Var=Σ(w/q)(Y−M/q)²=C/q³.Every term is nonnegative,so the sum is0 exactly when EVERY positive-weight Y equals the mean,equivalently constancy on positive support.Zero-weight values may differ.This is NOT n−1 sample variance.
Step identifiers and dependencies
Rule: Variance_CENTERED_definition_and_positive_support_zero_equivalence · Uses: G001, S037
S039 · implies
Before
source roles = [shift_variance, null, -6]
After
derived = [720/49]
Conditions: none
Under a shift,the mean also shifts by c,leaving deviations unchanged.Under scaling,the mean scales by k and squared deviations by k²,including negative/zero k.The ACTUAL Y values were fully recomputed;the law does not excuse untyped c/k.
Step identifiers and dependencies
Rule: Variance_SHIFT_SCALE_law_with_conditions · Uses: G001, S038