CBLUP90IOD2
CBLUP90IOD2 is a threshold-linear program for obtaining solutions to very large models. Solutions are obtained by iteration on data using the preconditioned conjugate iteration.
History
- Original linear model program BLUPF90 (I. Misztal)
- Rewrite to threshold-linear model program CBLUP90 (I. Misztal)
- Rewrite to multiple linear traits with computed thresholds and multiple linear traits(B. Auvray)
- Rewrite of BLUPF90 to iteration on data: BLUP90IOD (S. Tsuruta)
Summary
CBLUP90IOD2 supports models with one categorical trait and multiple linear traits. Thresholds are computed but none of the variances is. Memory requirements are about 56 bytes per equation.
The first trait must be categorical and not be missing.
Options
OPTION category 3
Number of categories (default 2)
OPTION conv_crit 1d-12
Convergence criterion (default 1d-12).
OPTION maxrounds 500
Maximum rounds (default 5000).
OPTION pcg_maxrounds 40
Set the maximum number of PCG rounds (default 40).
OPTION cont
Keep the last solutions and residuals for restart. When the program does not converge, you will have “solutions_rsd”. This option requires this file for restart.
OPTION cont 1
Restart the program from the last run using the previous solutions.
OPTION thresholds t1 t2 ... 0
Give thresholds for categorical traits (the last threshold = 0).
OPTION avgeps
Use the average eps / PCG rounds for convergence. With this option, solutions will be more stable.
OPTION SNP_file snp
Specify the SNP file name to use genotype data.
Example 1
parameter file: ex1.par
DATAFILE ex1.dat NUMBER_OF_TRAITS 3 NUMBER_OF_EFFECTS 4 OBSERVATION(S) 16 14 15 WEIGHT(S) EFFECTS: POSITIONS_IN_DATAFILE NUMBER_OF_LEVELS TYPE_OF_EFFECT [EFFECT NESTED] 1 1 1 53 cross 5 5 5 2 cross 6 6 6 1 cross 10 0 10 1 cross RANDOM_RESIDUAL VALUES 1 15 3 15 726 29.8509 3 29.8509 22.078 RANDOM_GROUP 1 RANDOM_TYPE add_animal FILE ex1.ped (CO)VARIANCES 0.42857 -7.202 .6462 -7.202 484 15.198 .6462 15.198 3.9 OPTION category 3
data file: ex1.dat
53 6 0 7 2 1 0 6 1 0 10 1 0 290.0 36.5 1 52 6 0 7 2 1 0 6 1 0 10 1 0 317.8 43.5 2 51 6 0 7 1 0 0 5 0 0 9 0 0 256.5 36.0 1 50 6 0 7 1 0 0 5 0 1 9 0 1 315.0 45.0 1 49 6 0 7 1 0 0 5 0 1 9 0 1 273.0 46.0 2 48 6 0 7 1 0 0 5 0 1 9 0 1 354.8 50.5 3 47 6 0 7 1 1 0 5 1 0 9 1 0 357.0 36.5 1 46 6 0 7 1 1 0 5 1 1 9 1 1 260.0 45.3 2 45 6 0 7 1 1 0 5 1 0 9 1 0 290.0 51.6 3 44 5 0 7 2 0 0 6 0 1 10 0 1 346.5 51.0 2 43 5 0 7 2 0 0 6 0 1 10 0 1 317.8 43.7 2 42 5 0 7 2 1 0 6 1 1 10 1 1 357.0 50.5 2 41 5 0 7 2 1 0 6 1 0 10 1 0 357.0 41.0 1 40 5 0 7 1 0 0 5 0 0 9 0 0 344.0 48.0 3 39 5 0 7 1 0 0 5 0 1 9 0 1 346.5 52.0 1 38 5 0 7 1 0 0 5 0 1 9 0 1 328.0 38.5 1 37 5 0 7 1 0 0 5 0 1 9 0 1 285.0 48.8 2 36 5 0 7 1 0 0 5 0 1 9 0 1 396.0 42.5 1 35 5 0 7 1 1 0 5 1 0 9 1 0 304.5 43.5 1 34 5 0 7 1 1 0 5 1 1 9 1 1 346.5 40.5 1 33 4 0 7 2 1 0 6 1 1 10 1 1 333.3 44.5 1 32 4 0 7 1 0 0 5 0 0 9 0 0 243.0 39.0 1 31 4 0 7 1 0 0 5 0 0 9 0 0 344.0 51.0 2 30 4 0 7 1 0 0 5 0 1 9 0 1 328.0 47.0 2 29 3 0 7 2 0 0 6 0 1 10 0 1 320.0 46.0 1 28 3 0 7 2 0 0 6 0 0 10 0 0 302.3 44.4 1 27 3 0 7 2 0 0 6 0 0 10 0 0 266.0 32.5 1 26 3 0 7 2 1 0 6 1 0 10 1 0 328.0 42.7 1 25 3 0 7 2 1 0 6 1 1 10 1 1 300.0 36.0 1 24 3 0 7 1 1 0 5 1 1 9 1 1 341.0 41.6 1 23 2 0 7 2 0 0 6 0 1 10 0 1 273.0 49.0 1 22 2 0 7 2 0 0 6 0 1 10 0 1 300.0 44.5 1 21 2 0 7 1 0 0 5 0 1 9 0 1 315.0 42.0 1 20 2 0 7 1 0 0 5 0 1 9 0 1 307.5 47.0 2 19 2 0 7 1 0 0 5 0 0 9 0 0 346.5 34.0 1 18 2 0 7 1 1 0 5 1 1 9 1 1 333.3 43.0 2 17 2 0 7 1 1 0 5 1 1 9 1 1 346.5 41.4 1 16 1 0 7 2 0 0 6 0 0 10 0 0 346.5 39.0 1 15 1 0 7 2 1 0 6 1 0 10 1 0 374.0 40.5 1 14 1 0 7 2 1 0 6 1 0 10 1 0 336.0 46.0 1 13 1 0 7 1 0 0 5 0 0 9 0 0 270.0 35.0 1 12 1 0 7 1 0 0 5 0 0 9 0 0 310.0 42.0 1 11 1 0 7 1 0 0 5 0 0 9 0 0 285.0 43.0 1 10 1 0 7 1 0 0 5 0 0 9 0 0 374.0 40.0 1 9 1 0 7 1 1 0 5 1 0 9 1 0 354.8 41.5 1 8 1 0 7 1 1 0 5 1 1 9 1 1 304.5 37.5 1 7 1 0 7 1 1 0 5 1 1 9 1 1 328.0 41.0 1
pedigree file: ex1.ped
53 6 0 52 6 0 51 6 0 50 6 0 49 6 0 48 6 0 47 6 0 46 6 0 45 6 0 44 5 0 43 5 0 42 5 0 41 5 0 40 5 0 39 5 0 38 5 0 37 5 0 36 5 0 35 5 0 34 5 0 33 4 0 32 4 0 31 4 0 30 4 0 29 3 0 28 3 0 27 3 0 26 3 0 25 3 0 24 3 0 23 2 0 22 2 0 21 2 0 20 2 0 19 2 0 18 2 0 17 2 0 16 1 0 15 1 0 14 1 0 13 1 0 12 1 0 11 1 0 10 1 0 9 1 0 8 1 0 7 1 0 6 0 0 5 0 0 4 0 0 3 0 0 2 0 0 1 0 0
Output log:
number of categories (numbered from 1)? name of parameter file?ex1.par CBLUP90IOD 1.0 by I. Misztal, 10/11/99 Parameter file: ex1.par Data file: ex1.dat Number of Traits 3 Number of Effects 4 Position of Observations 16 14 15 Position of Weight (1) 0 Value of Missing Trait/Observation 0 EFFECTS # type position (2) levels [positions for nested] 1 cross-classified 1 1 1 53 2 cross-classified 5 5 5 2 3 cross-classified 6 6 6 1 4 cross-classified 10 0 10 1 Residual (co)variance Matrix 1.0000 15.000 3.0000 15.000 726.00 29.851 3.0000 29.851 22.078 Random Effect(s) 1 Type of Random Effect: additive animal Pedigree File: ex1.ped trait effect (CO)VARIANCES 1 1 0.42857 -7.2020 0.64620 2 1 -7.2020 484.00 15.198 3 1 0.64620 15.198 3.9000 REMARKS (1) Weight position 0 means no weights utilized (2) Effect positions of 0 for some effects and traits means that such effects are missing for specified traits Data record length = 16 number of rounds? initial thresholds: -0.5000 -0.0000 0.42857 -7.2020 0.64620 -7.2020 484.00 15.198 0.64620 15.198 3.9000 G inverse 10.094 0.23098 -2.5726 0.23098 0.76397E-02 -0.68043E-01 -2.5726 -0.68043E-01 0.94782 read 47 records eps = 1.545122445227030E-011 no. of iteration = 49 thresholds in round 1: -0.47537 0.00000 eps = 2.114189614579900E-011 no. of iteration = 46 thresholds in round 2: -0.99853 0.00000 eps = 3.195444694991311E-011 no. of iteration = 44 thresholds in round 3: -1.21192 0.00000 eps = 4.229633202398320E-011 no. of iteration = 44 thresholds in round 4: -1.26689 0.00000 solutions stored in file: "solutions"
Example 2
parameter file: ex2.par
- Large simulated data set
- 3 traits, first one categorical (4 categories).
DATAFILE ex2.dat NUMBER_OF_TRAITS 3 NUMBER_OF_EFFECTS 2 OBSERVATION(S) 3 4 5 WEIGHT(S) EFFECTS: POSITIONS_IN_DATAFILE NUMBER_OF_LEVELS TYPE_OF_EFFECT [EFFECT NESTED] 2 2 2 50 cross 1 1 1 1100 cross RANDOM_RESIDUAL VALUES 1.0000 7.7460 7.7460 7.7460 1351.0 231.81 7.7460 231.81 1369.0 RANDOM_GROUP 2 RANDOM_TYPE add_animal FILE ex2.ped (CO)VARIANCES 0.66667 5.16398 5.16398 5.16398 1000.00000 200.00000 5.16398 200.00000 1000.00000
Output log:
number of categories (numbered from 1)? name of parameter file?ex2.par CBLUP90IOD 1.0 by I. Misztal, 10/11/99 Parameter file: ex2.par Data file: ex2.dat Number of Traits 3 Number of Effects 2 Position of Observations 3 4 5 Position of Weight (1) 0 Value of Missing Trait/Observation 0 EFFECTS # type position (2) levels [positions for nested] 1 cross-classified 2 2 2 50 2 cross-classified 1 1 1 1100 Residual (co)variance Matrix 1.0000 7.7460 7.7460 7.7460 1351.0 231.81 7.7460 231.81 1369.0 Random Effect(s) 2 Type of Random Effect: additive animal Pedigree File: ex2.ped trait effect (CO)VARIANCES 1 2 0.66667 5.1640 5.1640 2 2 5.1640 1000.0 200.00 3 2 5.1640 200.00 1000.0 REMARKS (1) Weight position 0 means no weights utilized (2) Effect positions of 0 for some effects and traits means that such effects are missing for specified traits Data record length = 5 number of rounds? initial thresholds: -0.6667 -0.3333 -0.0000 0.66667 5.1640 5.1640 5.1640 1000.0 200.00 5.1640 200.00 1000.0 G inverse 1.6071 -0.69160E-02 -0.69160E-02 -0.69160E-02 0.10714E-02 -0.17857E-03 -0.69160E-02 -0.17857E-03 0.10714E-02 read 1000 records eps = 9.400106129966520E-011 no. of iteration = 41 thresholds in round 1: -1.38764 -0.61188 0.00000 eps = 9.044560960701230E-011 no. of iteration = 44 thresholds in round 2: -1.86832 -0.90333 0.00000 eps = 5.262946997597152E-011 no. of iteration = 46 thresholds in round 3: -2.08856 -1.02189 0.00000 eps = 8.666286752404070E-011 no. of iteration = 46 thresholds in round 4: -2.15157 -1.05558 0.00000 eps = 9.584502638477410E-011 no. of iteration = 46 thresholds in round 5: -2.16738 -1.06370 0.00000 solutions stored in file: "solutions"