bc_pa ~ outcome + absPE*PE_flag + ( 1 + outcome + absPE*PE_flag | cohort / id / exam_num
## Warning in commonArgs(par, fn, control, environment()): maxfun < 10 *
## length(par)^2 is not recommended.## Warning in optwrap(optimizer, devfun, getStart(start, rho$lower, rho$pp), :
## convergence code 1 from bobyqa: bobyqa -- maximum number of function
## evaluations exceeded## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : unable to evaluate scaled gradient## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : Model failed to converge: degenerate Hessian with 4 negative
## eigenvalues## Warning: Model failed to converge with 4 negative eigenvalues: -4.7e-01
## -5.0e-01 -2.0e+00 -2.0e+00## Linear mixed model fit by maximum likelihood . t-tests use
## Satterthwaite's method [lmerModLmerTest]
## Formula:
## bc_pa ~ outcome + absPE * PE_flag + (1 + outcome + absPE * PE_flag |
## cohort/id/exam_num)
## Data: df.pa.conf.new
##
## AIC BIC logLik deviance df.resid
## 22303.4 22607.0 -11100.7 22201.4 2794
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.5204 -0.4473 0.0252 0.4561 5.0624
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## exam_num:(id:cohort) (Intercept) 425.99760 20.6397
## outcome 2.90935 1.7057 -0.82
## absPE 19.00141 4.3591 -0.89 0.99
## PE_flag 212.09376 14.5634 -0.72 0.51 0.57
## absPE:PE_flag 98.99247 9.9495 0.66 -0.91 -0.88
## id:cohort (Intercept) 61.95545 7.8712
## outcome 1.03110 1.0154 0.02
## absPE 9.82848 3.1350 -1.00 -0.05
## PE_flag 81.94049 9.0521 -0.54 -0.83 0.56
## absPE:PE_flag 16.76150 4.0941 0.97 -0.16 -0.97
## cohort (Intercept) 8.94317 2.9905
## outcome 0.07967 0.2823 -1.00
## absPE 0.33612 0.5798 -1.00 1.00
## PE_flag 5.62046 2.3708 -1.00 1.00 1.00
## absPE:PE_flag 1.25797 1.1216 1.00 -1.00 -1.00
## Residual 94.14555 9.7029
##
##
##
##
##
## -0.59
##
##
##
##
## -0.41
##
##
##
##
## -1.00
##
## Number of obs: 2845, groups:
## exam_num:(id:cohort), 453; id:cohort, 156; cohort, 3
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) -3.2235 2.6161 3.7275 -1.232 0.28990
## outcome 0.2971 0.2460 3.5888 1.208 0.30063
## absPE -2.8597 0.8460 7.4107 -3.380 0.01079 *
## PE_flag 2.1393 2.4702 4.7754 0.866 0.42784
## absPE:PE_flag 4.9014 1.2828 5.6308 3.821 0.00987 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) outcom absPE PE_flg
## outcome -0.663
## absPE -0.756 0.270
## PE_flag -0.828 0.299 0.721
## absPE:PE_fl 0.764 -0.564 -0.769 -0.753
## convergence code: 1
## unable to evaluate scaled gradient
## Model failed to converge: degenerate Hessian with 4 negative eigenvalues
## maxfun < 10 * length(par)^2 is not recommended.bc_na ~ outcome + absPE*PE_flag + ( 1 + outcome + absPE*PE_flag | cohort / id / exam_num
## Warning in commonArgs(par, fn, control, environment()): maxfun < 10 *
## length(par)^2 is not recommended.## Warning in optwrap(optimizer, devfun, getStart(start, rho$lower, rho$pp), :
## convergence code 1 from bobyqa: bobyqa -- maximum number of function
## evaluations exceeded## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : unable to evaluate scaled gradient## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control
## $checkConv, : Model failed to converge: degenerate Hessian with 3 negative
## eigenvalues## Warning: Model failed to converge with 3 negative eigenvalues: -9.0e-01
## -2.4e+00 -3.4e+00## Linear mixed model fit by maximum likelihood . t-tests use
## Satterthwaite's method [lmerModLmerTest]
## Formula:
## bc_na ~ outcome + absPE * PE_flag + (1 + outcome + absPE * PE_flag |
## cohort/id/exam_num)
## Data: df.pa.conf.new
##
## AIC BIC logLik deviance df.resid
## 22800.0 23103.6 -11349.0 22698.0 2794
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.8496 -0.4772 -0.0333 0.4150 4.8073
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## exam_num:(id:cohort) (Intercept) 428.20227 20.6930
## outcome 5.09196 2.2565 -0.72
## absPE 10.87861 3.2983 -0.77 0.96
## PE_flag 53.42291 7.3091 -0.22 -0.26 -0.12
## absPE:PE_flag 37.03820 6.0859 0.37 -0.86 -0.86
## id:cohort (Intercept) 133.84152 11.5690
## outcome 0.97488 0.9874 -0.93
## absPE 11.68266 3.4180 -0.90 0.68
## PE_flag 25.70105 5.0696 -0.81 0.83 0.64
## absPE:PE_flag 48.45568 6.9610 0.96 -0.86 -0.90
## cohort (Intercept) 3.84233 1.9602
## outcome 0.02977 0.1726 -1.00
## absPE 0.43506 0.6596 -0.99 0.99
## PE_flag 10.10563 3.1789 -0.88 0.88 0.87
## absPE:PE_flag 1.82452 1.3507 0.96 -0.96 -0.96
## Residual 108.53625 10.4181
##
##
##
##
##
## 0.39
##
##
##
##
## -0.90
##
##
##
##
## -0.97
##
## Number of obs: 2845, groups:
## exam_num:(id:cohort), 453; id:cohort, 156; cohort, 3
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 6.8578 2.4228 5.7494 2.830 0.0314 *
## outcome -0.6367 0.2288 7.5423 -2.783 0.0252 *
## absPE 3.0183 0.9261 4.8287 3.259 0.0236 *
## PE_flag -2.1582 2.8115 4.7766 -0.768 0.4789
## absPE:PE_flag -4.5732 1.4531 4.1430 -3.147 0.0330 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) outcom absPE PE_flg
## outcome -0.629
## absPE -0.715 0.215
## PE_flag -0.722 0.223 0.653
## absPE:PE_fl 0.731 -0.523 -0.768 -0.751
## convergence code: 1
## unable to evaluate scaled gradient
## Model failed to converge: degenerate Hessian with 3 negative eigenvalues
## maxfun < 10 * length(par)^2 is not recommended.Affect ~ Affect_flag*(outcome + absPE*PE_flag) + ( 1 | cohort / id / exam_num)
## Linear mixed model fit by maximum likelihood . t-tests use
## Satterthwaite's method [lmerModLmerTest]
## Formula:
## Affect ~ Affect_flag * (outcome + absPE * PE_flag) + (1 | cohort/id/exam_num)
## Data: longDF
##
## AIC BIC logLik deviance df.resid
## 47755.3 47848.4 -23863.7 47727.3 5676
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -4.4330 -0.6199 -0.0174 0.5816 4.5795
##
## Random effects:
## Groups Name Variance Std.Dev.
## exam_num:(id:cohort) (Intercept) 2.98771 1.72850
## id:cohort (Intercept) 1.24212 1.11450
## cohort (Intercept) 0.00985 0.09925
## Residual 253.50962 15.92199
## Number of obs: 5690, groups:
## exam_num:(id:cohort), 453; id:cohort, 156; cohort, 3
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 10.2840 1.2443 733.7137 8.265 6.55e-16
## Affect_flag -15.5186 1.6745 5342.5112 -9.268 < 2e-16
## outcome -0.9298 0.1423 1808.6651 -6.536 8.20e-11
## absPE 1.7062 0.3535 1702.4319 4.826 1.51e-06
## PE_flag -3.4453 1.0122 1405.2468 -3.404 0.000683
## absPE:PE_flag -2.3282 0.5454 1562.3794 -4.269 2.08e-05
## Affect_flag:outcome 1.3282 0.1923 5342.5112 6.907 5.52e-12
## Affect_flag:absPE -3.2340 0.4753 5342.5112 -6.804 1.13e-11
## Affect_flag:PE_flag 7.2601 1.3522 5342.5112 5.369 8.25e-08
## Affect_flag:absPE:PE_flag 4.8869 0.7300 5342.5112 6.695 2.38e-11
##
## (Intercept) ***
## Affect_flag ***
## outcome ***
## absPE ***
## PE_flag ***
## absPE:PE_flag ***
## Affect_flag:outcome ***
## Affect_flag:absPE ***
## Affect_flag:PE_flag ***
## Affect_flag:absPE:PE_flag ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Affct_ outcom absPE PE_flg aPE:PE Affc_: Af_:PE A_:PE_
## Affect_flag -0.673
## outcome -0.766 0.523
## absPE -0.695 0.475 0.309
## PE_flag -0.423 0.280 -0.100 0.529
## absPE:PE_fl 0.552 -0.374 -0.335 -0.688 -0.661
## Affct_flg:t 0.521 -0.774 -0.676 -0.218 0.068 0.227
## Affct_fl:PE 0.475 -0.706 -0.219 -0.672 -0.355 0.463 0.324
## Affct_f:PE_ 0.282 -0.419 0.068 -0.357 -0.668 0.445 -0.101 0.532
## Aff_:PE:PE_ -0.376 0.559 0.229 0.466 0.445 -0.669 -0.339 -0.693 -0.666Note: - No random effects in model. Unsure of how to specify random effects with a nested interaction. - Model fails to converge when nested interaction is included in the random effects