tests/testthat/_snaps/proportional_hazards-glmnet.md

In EmilHvitfeldt/survnip: 'parsnip' Engines for Survival Models

print coxnet model

Code
  f_fit
Output
  parsnip model object


  Call:  glmnet::glmnet(x = data_obj$x, y = data_obj$y, family = "cox",      weights = weights, alpha = alpha, lambda = lambda)

     Df %Dev   Lambda
  1   0 0.00 0.225300
  2   1 0.21 0.205300
  3   1 0.39 0.187100
  4   1 0.53 0.170500
  5   1 0.65 0.155300
  6   1 0.75 0.141500
  7   1 0.84 0.128900
  8   1 0.90 0.117500
  9   1 0.96 0.107000
  10  1 1.01 0.097540
  11  1 1.05 0.088870
  12  2 1.08 0.080980
  13  2 1.13 0.073790
  14  2 1.16 0.067230
  15  2 1.19 0.061260
  16  2 1.22 0.055820
  17  2 1.24 0.050860
  18  2 1.26 0.046340
  19  2 1.27 0.042220
  20  2 1.28 0.038470
  21  2 1.29 0.035050
  22  2 1.30 0.031940
  23  2 1.31 0.029100
  24  2 1.31 0.026520
  25  2 1.32 0.024160
  26  2 1.32 0.022010
  27  2 1.33 0.020060
  28  2 1.33 0.018280
  29  2 1.33 0.016650
  30  2 1.33 0.015170
  31  2 1.33 0.013830
  32  2 1.33 0.012600
  33  2 1.34 0.011480
  34  2 1.34 0.010460
  35  2 1.34 0.009530
  36  2 1.34 0.008683
  37  2 1.34 0.007912
  38  2 1.34 0.007209
  39  2 1.34 0.006568
  40  2 1.34 0.005985
  41  2 1.34 0.005453
  The training data has been saved for prediction.

stratification is specified in a single term

Code
  fit(spec, Surv(time, status) ~ age + ph.ecog + strata(sex) + strata(inst),
  data = lung)
Condition
  Error:
  ! There can only be a single strata term specified using the `strata()` function.
  i It can contain multiple strata columns, e.g., `~ x + strata(s1, s2)`.

formula modifications to remove strata

Code
  fit(spec, Surv(time, status) ~ strata(sex), data = lung)
Condition
  Error:
  ! The Cox model does not contain an intercept, please add a predictor.

Code
  fit(spec, Surv(time, status) ~ age + (ph.ecog + strata(sex)), data = lung)
Condition
  Error:
  ! Stratification must be nested under a chain of `+` calls.
  i # Good: `~ x1 + x2 + strata(s)`
  i # Bad: `~ x1 + (x2 + strata(s))`

protect certain glmnet engine args

Code
  proportional_hazards(penalty = 0.1) %>% set_engine("glmnet", family = "gaussian") %>%
    fit(Surv(time, status) ~ age + sex, data = lung)
Condition
  Error:
  ! This argument cannot be used to create the model: `family`.

predictions with strata and dot in formula

Code
  f_fit <- fit(cox_spec, Surv(time, status) ~ . - sex + strata(sex), data = lung2)
Condition
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge

Code
  f_fit_2 <- fit(cox_spec, Surv(time, status) ~ ph.ecog + age + strata(sex),
  data = lung2)
Condition
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge
  Warning:
  cox.fit: algorithm did not converge

Code
  predict(f_fit, lung2, type = "linear_pred")
Output
  # A tibble: 227 x 1
     .pred_linear_pred
                 <dbl>
   1            -1.09 
   2            -0.579
   3            -0.477
   4            -0.940
   5            -0.511
   6            -1.09 
   7            -1.49 
   8            -1.51 
   9            -0.906
  10            -1.43 
  # i 217 more rows
Code
  predict(f_fit, lung2, type = "survival", eval_time = c(100, 300))
Condition
  Warning in `terms.formula()`:
  'varlist' has changed (from nvar=5) to new 6 after EncodeVars() -- should no longer happen!
Output
  # A tibble: 227 x 1
     .pred           
     <list>          
   1 <tibble [2 x 2]>
   2 <tibble [2 x 2]>
   3 <tibble [2 x 2]>
   4 <tibble [2 x 2]>
   5 <tibble [2 x 2]>
   6 <tibble [2 x 2]>
   7 <tibble [2 x 2]>
   8 <tibble [2 x 2]>
   9 <tibble [2 x 2]>
  10 <tibble [2 x 2]>
  # i 217 more rows

Code
  f_pred <- predict(f_fit, lung2, type = "survival", eval_time = c(100, 300))
Condition
  Warning in `terms.formula()`:
  'varlist' has changed (from nvar=5) to new 6 after EncodeVars() -- should no longer happen!
Code
  f_pred_2 <- predict(f_fit_2, lung2, type = "survival", eval_time = c(100, 300))

fit_xy() errors with stratification

Code
  fit_xy(spec, x = lung_x, y = lung_y_s)
Condition
  Error in `fit_xy()`:
  ! For stratification, please use the formula interface via `fit()`.

EmilHvitfeldt/survnip documentation built on April 23, 2024, 3:07 p.m.