Modeling censored knowledge with tfprobability

Nothing’s ever good, and knowledge isn’t both. One sort of “imperfection” is lacking knowledge, the place some options are unobserved for some topics. (A subject for one more submit.) One other is censored knowledge, the place an occasion whose traits we need to measure doesn’t happen within the remark interval. The instance in Richard McElreath’s Statistical Rethinking is time to adoption of cats in an animal shelter. If we repair an interval and observe wait occasions for these cats that truly did get adopted, our estimate will find yourself too optimistic: We don’t keep in mind these cats who weren’t adopted throughout this interval and thus, would have contributed wait occasions of size longer than the whole interval.

On this submit, we use a barely much less emotional instance which nonetheless could also be of curiosity, particularly to R bundle builders: time to completion of R CMD verify, collected from CRAN and offered by the parsnip bundle as check_times. Right here, the censored portion are these checks that errored out for no matter cause, i.e., for which the verify didn’t full.

Why can we care concerning the censored portion? Within the cat adoption situation, that is fairly apparent: We wish to have the ability to get a practical estimate for any unknown cat, not simply these cats that can develop into “fortunate”. How about check_times? Effectively, in case your submission is a type of that errored out, you continue to care about how lengthy you wait, so despite the fact that their proportion is low (< 1%) we don’t need to merely exclude them. Additionally, there may be the chance that the failing ones would have taken longer, had they run to completion, as a result of some intrinsic distinction between each teams. Conversely, if failures have been random, the longer-running checks would have a better likelihood to get hit by an error. So right here too, exluding the censored knowledge might end in bias.

How can we mannequin durations for that censored portion, the place the “true length” is unknown? Taking one step again, how can we mannequin durations usually? Making as few assumptions as attainable, the most entropy distribution for displacements (in house or time) is the exponential. Thus, for the checks that truly did full, durations are assumed to be exponentially distributed.

For the others, all we all know is that in a digital world the place the verify accomplished, it could take at the least as lengthy because the given length. This amount will be modeled by the exponential complementary cumulative distribution perform (CCDF). Why? A cumulative distribution perform (CDF) signifies the chance {that a} worth decrease or equal to some reference level was reached; e.g., “the chance of durations <= 255 is 0.9”. Its complement, 1 – CDF, then offers the chance {that a} worth will exceed than that reference level.

Let’s see this in motion.

The info

The next code works with the present steady releases of TensorFlow and TensorFlow Likelihood, that are 1.14 and 0.7, respectively. In the event you don’t have tfprobability put in, get it from Github:

These are the libraries we want. As of TensorFlow 1.14, we name tf$compat$v2$enable_v2_behavior() to run with keen execution.

Moreover the verify durations we need to mannequin, check_times reviews numerous options of the bundle in query, akin to variety of imported packages, variety of dependencies, dimension of code and documentation information, and many others. The standing variable signifies whether or not the verify accomplished or errored out.

df <- check_times %>% choose(-bundle)
glimpse(df)

Observations: 13,626
Variables: 24
$ authors        <int> 1, 1, 1, 1, 5, 3, 2, 1, 4, 6, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1,…
$ imports        <dbl> 0, 6, 0, 0, 3, 1, 0, 4, 0, 7, 0, 0, 0, 0, 3, 2, 14, 2, 2, 0…
$ suggests       <dbl> 2, 4, 0, 0, 2, 0, 2, 2, 0, 0, 2, 8, 0, 0, 2, 0, 1, 3, 0, 0,…
$ relies upon        <dbl> 3, 1, 6, 1, 1, 1, 5, 0, 1, 1, 6, 5, 0, 0, 0, 1, 1, 5, 0, 2,…
$ Roxygen        <dbl> 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 0,…
$ gh             <dbl> 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0,…
$ rforge         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ descr          <int> 217, 313, 269, 63, 223, 1031, 135, 344, 204, 335, 104, 163,…
$ r_count        <int> 2, 20, 8, 0, 10, 10, 16, 3, 6, 14, 16, 4, 1, 1, 11, 5, 7, 1…
$ r_size         <dbl> 0.029053, 0.046336, 0.078374, 0.000000, 0.019080, 0.032607,…
$ ns_import      <dbl> 3, 15, 6, 0, 4, 5, 0, 4, 2, 10, 5, 6, 1, 0, 2, 2, 1, 11, 0,…
$ ns_export      <dbl> 0, 19, 0, 0, 10, 0, 0, 2, 0, 9, 3, 4, 0, 1, 10, 0, 16, 0, 2…
$ s3_methods     <dbl> 3, 0, 11, 0, 0, 0, 0, 2, 0, 23, 0, 0, 2, 5, 0, 4, 0, 0, 0, …
$ s4_methods     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,…
$ doc_count      <int> 0, 3, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,…
$ doc_size       <dbl> 0.000000, 0.019757, 0.038281, 0.000000, 0.007874, 0.000000,…
$ src_count      <int> 0, 0, 0, 0, 0, 0, 0, 2, 0, 5, 3, 0, 0, 0, 0, 0, 0, 54, 0, 0…
$ src_size       <dbl> 0.000000, 0.000000, 0.000000, 0.000000, 0.000000, 0.000000,…
$ data_count     <int> 2, 0, 0, 3, 3, 1, 10, 0, 4, 2, 2, 146, 0, 0, 0, 0, 0, 10, 0…
$ data_size      <dbl> 0.025292, 0.000000, 0.000000, 4.885864, 4.595504, 0.006500,…
$ testthat_count <int> 0, 8, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 3, 3, 0, 0,…
$ testthat_size  <dbl> 0.000000, 0.002496, 0.000000, 0.000000, 0.000000, 0.000000,…
$ check_time     <dbl> 49, 101, 292, 21, 103, 46, 78, 91, 47, 196, 200, 169, 45, 2…
$ standing         <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…

Of those 13,626 observations, simply 103 are censored:

0     1 
103 13523 

For higher readability, we’ll work with a subset of the columns. We use surv_reg to assist us discover a helpful and fascinating subset of predictors:

survreg_fit <-
  surv_reg(dist = "exponential") %>% 
  set_engine("survreg") %>% 
  match(Surv(check_time, standing) ~ ., 
      knowledge = df)
tidy(survreg_fit) 

# A tibble: 23 x 7
   time period             estimate std.error statistic  p.worth conf.low conf.excessive
   <chr>               <dbl>     <dbl>     <dbl>    <dbl>    <dbl>     <dbl>
 1 (Intercept)     3.86      0.0219     176.     0.             NA        NA
 2 authors         0.0139    0.00580      2.40   1.65e- 2       NA        NA
 3 imports         0.0606    0.00290     20.9    7.49e-97       NA        NA
 4 suggests        0.0332    0.00358      9.28   1.73e-20       NA        NA
 5 relies upon         0.118     0.00617     19.1    5.66e-81       NA        NA
 6 Roxygen         0.0702    0.0209       3.36   7.87e- 4       NA        NA
 7 gh              0.00898   0.0217       0.414  6.79e- 1       NA        NA
 8 rforge          0.0232    0.0662       0.351  7.26e- 1       NA        NA
 9 descr           0.000138  0.0000337    4.10   4.18e- 5       NA        NA
10 r_count         0.00209   0.000525     3.98   7.03e- 5       NA        NA
11 r_size          0.481     0.0819       5.87   4.28e- 9       NA        NA
12 ns_import       0.00352   0.000896     3.93   8.48e- 5       NA        NA
13 ns_export      -0.00161   0.000308    -5.24   1.57e- 7       NA        NA
14 s3_methods      0.000449  0.000421     1.06   2.87e- 1       NA        NA
15 s4_methods     -0.00154   0.00206     -0.745  4.56e- 1       NA        NA
16 doc_count       0.0739    0.0117       6.33   2.44e-10       NA        NA
17 doc_size        2.86      0.517        5.54   3.08e- 8       NA        NA
18 src_count       0.0122    0.00127      9.58   9.96e-22       NA        NA
19 src_size       -0.0242    0.0181      -1.34   1.82e- 1       NA        NA
20 data_count      0.0000415 0.000980     0.0423 9.66e- 1       NA        NA
21 data_size       0.0217    0.0135       1.61   1.08e- 1       NA        NA
22 testthat_count -0.000128  0.00127     -0.101  9.20e- 1       NA        NA
23 testthat_size   0.0108    0.0139       0.774  4.39e- 1       NA        NA

Evidently if we select imports, relies upon, r_size, doc_size, ns_import and ns_export we find yourself with a mixture of (comparatively) highly effective predictors from completely different semantic areas and of various scales.

Earlier than pruning the dataframe, we save away the goal variable. In our mannequin and coaching setup, it’s handy to have censored and uncensored knowledge saved individually, so right here we create two goal matrices as an alternative of 1:

# verify occasions for failed checks
# _c stands for censored
check_time_c <- df %>%
  filter(standing == 0) %>%
  choose(check_time) %>%
  as.matrix()

# verify occasions for profitable checks 
check_time_nc <- df %>%
  filter(standing == 1) %>%
  choose(check_time) %>%
  as.matrix()

Now we are able to zoom in on the variables of curiosity, organising one dataframe for the censored knowledge and one for the uncensored knowledge every. All predictors are normalized to keep away from overflow throughout sampling. We add a column of 1s to be used as an intercept.

df <- df %>% choose(standing,
                    relies upon,
                    imports,
                    doc_size,
                    r_size,
                    ns_import,
                    ns_export) %>%
  mutate_at(.vars = 2:7, .funs = perform(x) (x - min(x))/(max(x)-min(x))) %>%
  add_column(intercept = rep(1, nrow(df)), .earlier than = 1)

# dataframe of predictors for censored knowledge  
df_c <- df %>% filter(standing == 0) %>% choose(-standing)
# dataframe of predictors for non-censored knowledge 
df_nc <- df %>% filter(standing == 1) %>% choose(-standing)

That’s it for preparations. However after all we’re curious. Do verify occasions look completely different? Do predictors – those we selected – look completely different?

Evaluating a number of significant percentiles for each courses, we see that durations for uncompleted checks are increased than these for accomplished checks all through, aside from the 100% percentile. It’s not stunning that given the large distinction in pattern dimension, most length is increased for accomplished checks. In any other case although, doesn’t it seem like the errored-out bundle checks “have been going to take longer”?

How concerning the predictors? We don’t see any variations for relies upon, the variety of bundle dependencies (aside from, once more, the upper most reached for packages whose verify accomplished):

However for all others, we see the identical sample as reported above for check_time. Variety of packages imported is increased for censored knowledge in any respect percentiles apart from the utmost:

Similar for ns_export, the estimated variety of exported capabilities or strategies:

In addition to for ns_import, the estimated variety of imported capabilities or strategies:

Similar sample for r_size, the scale on disk of information within the R listing:

And eventually, we see it for doc_size too, the place doc_size is the scale of .Rmd and .Rnw information:

Given our process at hand – mannequin verify durations taking into consideration uncensored in addition to censored knowledge – we gained’t dwell on variations between each teams any longer; nonetheless we thought it fascinating to narrate these numbers.

So now, again to work. We have to create a mannequin.

The mannequin

As defined within the introduction, for accomplished checks length is modeled utilizing an exponential PDF. That is as simple as including tfd_exponential() to the mannequin perform, tfd_joint_distribution_sequential(). For the censored portion, we want the exponential CCDF. This one isn’t, as of as we speak, simply added to the mannequin. What we are able to do although is calculate its worth ourselves and add it to the “major” mannequin probability. We’ll see this under when discussing sampling; for now it means the mannequin definition finally ends up simple because it solely covers the non-censored knowledge. It’s fabricated from simply the stated exponential PDF and priors for the regression parameters.

As for the latter, we use 0-centered, Gaussian priors for all parameters. Normal deviations of 1 turned out to work effectively. Because the priors are all the identical, as an alternative of itemizing a bunch of tfd_normals, we are able to create them suddenly as

tfd_sample_distribution(tfd_normal(0, 1), sample_shape = 7)

Imply verify time is modeled as an affine mixture of the six predictors and the intercept. Right here then is the whole mannequin, instantiated utilizing the uncensored knowledge solely:

mannequin <- perform(knowledge) {
  tfd_joint_distribution_sequential(
    listing(
      tfd_sample_distribution(tfd_normal(0, 1), sample_shape = 7),
      perform(betas)
        tfd_independent(
          tfd_exponential(
            price = 1 / tf$math$exp(tf$transpose(
              tf$matmul(tf$solid(knowledge, betas$dtype), tf$transpose(betas))))),
          reinterpreted_batch_ndims = 1)))
}

m <- mannequin(df_nc %>% as.matrix())

All the time, we take a look at if samples from that mannequin have the anticipated shapes:

samples <- m %>% tfd_sample(2)
samples

[[1]]
tf.Tensor(
[[ 1.4184642   0.17583323 -0.06547955 -0.2512014   0.1862184  -1.2662812
   1.0231884 ]
 [-0.52142304 -1.0036682   2.2664437   1.29737     1.1123234   0.3810004
   0.1663677 ]], form=(2, 7), dtype=float32)

[[2]]
tf.Tensor(
[[4.4954767  7.865639   1.8388556  ... 7.914391   2.8485563  3.859719  ]
 [1.549662   0.77833986 0.10015647 ... 0.40323067 3.42171    0.69368565]], form=(2, 13523), dtype=float32)

This seems to be effective: We’ve an inventory of size two, one factor for every distribution within the mannequin. For each tensors, dimension 1 displays the batch dimension (which we arbitrarily set to 2 on this take a look at), whereas dimension 2 is 7 for the variety of regular priors and 13523 for the variety of durations predicted.

How doubtless are these samples?

m %>% tfd_log_prob(samples)

tf.Tensor([-32464.521   -7693.4023], form=(2,), dtype=float32)

Right here too, the form is appropriate, and the values look cheap.

The following factor to do is outline the goal we need to optimize.

Optimization goal

Abstractly, the factor to maximise is the log probility of the information – that’s, the measured durations – below the mannequin.
Now right here the information is available in two elements, and the goal does as effectively. First, we now have the non-censored knowledge, for which

m %>% tfd_log_prob(listing(betas, tf$solid(target_nc, betas$dtype)))

will calculate the log chance. Second, to acquire log chance for the censored knowledge we write a customized perform that calculates the log of the exponential CCDF:

get_exponential_lccdf <- perform(betas, knowledge, goal) {
  e <-  tfd_independent(tfd_exponential(price = 1 / tf$math$exp(tf$transpose(tf$matmul(
    tf$solid(knowledge, betas$dtype), tf$transpose(betas)
  )))),
  reinterpreted_batch_ndims = 1)
  cum_prob <- e %>% tfd_cdf(tf$solid(goal, betas$dtype))
  tf$math$log(1 - cum_prob)
}

Each elements are mixed in somewhat wrapper perform that permits us to match coaching together with and excluding the censored knowledge. We gained’t try this on this submit, however you may be to do it with your personal knowledge, particularly if the ratio of censored and uncensored elements is rather less imbalanced.

get_log_prob <-
  perform(target_nc,
           censored_data = NULL,
           target_c = NULL) {
    log_prob <- perform(betas) {
      log_prob <-
        m %>% tfd_log_prob(listing(betas, tf$solid(target_nc, betas$dtype)))
      potential <-
        if (!is.null(censored_data) && !is.null(target_c))
          get_exponential_lccdf(betas, censored_data, target_c)
      else
        0
      log_prob + potential
    }
    log_prob
  }

log_prob <-
  get_log_prob(
    check_time_nc %>% tf$transpose(),
    df_c %>% as.matrix(),
    check_time_c %>% tf$transpose()
  )

Sampling

With mannequin and goal outlined, we’re able to do sampling.

n_chains <- 4
n_burnin <- 1000
n_steps <- 1000

# preserve observe of some diagnostic output, acceptance and step dimension
trace_fn <- perform(state, pkr) {
  listing(
    pkr$inner_results$is_accepted,
    pkr$inner_results$accepted_results$step_size
  )
}

# get form of preliminary values 
# to start out sampling with out producing NaNs, we'll feed the algorithm
# tf$zeros_like(initial_betas)
# as an alternative 
initial_betas <- (m %>% tfd_sample(n_chains))[[1]]

For the variety of leapfrog steps and the step dimension, experimentation confirmed {that a} mixture of 64 / 0.1 yielded cheap outcomes:

hmc <- mcmc_hamiltonian_monte_carlo(
  target_log_prob_fn = log_prob,
  num_leapfrog_steps = 64,
  step_size = 0.1
) %>%
  mcmc_simple_step_size_adaptation(target_accept_prob = 0.8,
                                   num_adaptation_steps = n_burnin)

run_mcmc <- perform(kernel) {
  kernel %>% mcmc_sample_chain(
    num_results = n_steps,
    num_burnin_steps = n_burnin,
    current_state = tf$ones_like(initial_betas),
    trace_fn = trace_fn
  )
}

# vital for efficiency: run HMC in graph mode
run_mcmc <- tf_function(run_mcmc)

res <- hmc %>% run_mcmc()
samples <- res$all_states

Outcomes

Earlier than we examine the chains, here’s a fast take a look at the proportion of accepted steps and the per-parameter imply step dimension:

0.995

0.004953894

We additionally retailer away efficient pattern sizes and the rhat metrics for later addition to the synopsis.

effective_sample_size <- mcmc_effective_sample_size(samples) %>%
  as.matrix() %>%
  apply(2, imply)
potential_scale_reduction <- mcmc_potential_scale_reduction(samples) %>%
  as.numeric()

We then convert the samples tensor to an R array to be used in postprocessing.

# 2-item listing, the place every merchandise has dim (1000, 4)
samples <- as.array(samples) %>% array_branch(margin = 3)

How effectively did the sampling work? The chains combine effectively, however for some parameters, autocorrelation remains to be fairly excessive.

prep_tibble <- perform(samples) {
  as_tibble(samples,
            .name_repair = ~ c("chain_1", "chain_2", "chain_3", "chain_4")) %>%
    add_column(pattern = 1:n_steps) %>%
    collect(key = "chain", worth = "worth",-pattern)
}

plot_trace <- perform(samples) {
  prep_tibble(samples) %>%
    ggplot(aes(x = pattern, y = worth, coloration = chain)) +
    geom_line() +
    theme_light() +
    theme(
      legend.place = "none",
      axis.title = element_blank(),
      axis.textual content = element_blank(),
      axis.ticks = element_blank()
    )
}

plot_traces <- perform(samples) {
  plots <- purrr::map(samples, plot_trace)
  do.name(grid.organize, plots)
}

plot_traces(samples)

Determine 1: Hint plots for the 7 parameters.

Now for a synopsis of posterior parameter statistics, together with the same old per-parameter sampling indicators efficient pattern dimension and rhat.

all_samples <- map(samples, as.vector)

means <- map_dbl(all_samples, imply)

sds <- map_dbl(all_samples, sd)

hpdis <- map(all_samples, ~ hdi(.x) %>% t() %>% as_tibble())

abstract <- tibble(
  imply = means,
  sd = sds,
  hpdi = hpdis
) %>% unnest() %>%
  add_column(param = colnames(df_c), .after = FALSE) %>%
  add_column(
    n_effective = effective_sample_size,
    rhat = potential_scale_reduction
  )

abstract

# A tibble: 7 x 7
  param       imply     sd  decrease higher n_effective  rhat
  <chr>      <dbl>  <dbl>  <dbl> <dbl>       <dbl> <dbl>
1 intercept  4.05  0.0158  4.02   4.08       508.   1.17
2 relies upon    1.34  0.0732  1.18   1.47      1000    1.00
3 imports    2.89  0.121   2.65   3.12      1000    1.00
4 doc_size   6.18  0.394   5.40   6.94       177.   1.01
5 r_size     2.93  0.266   2.42   3.46       289.   1.00
6 ns_import  1.54  0.274   0.987  2.06       387.   1.00
7 ns_export -0.237 0.675  -1.53   1.10        66.8  1.01

Determine 2: Posterior means and HPDIs.

From the diagnostics and hint plots, the mannequin appears to work fairly effectively, however as there is no such thing as a simple error metric concerned, it’s exhausting to know if precise predictions would even land in an applicable vary.

To ensure they do, we examine predictions from our mannequin in addition to from surv_reg.
This time, we additionally break up the information into coaching and take a look at units. Right here first are the predictions from surv_reg:

train_test_split <- initial_split(check_times, strata = "standing")
check_time_train <- coaching(train_test_split)
check_time_test <- testing(train_test_split)

survreg_fit <-
  surv_reg(dist = "exponential") %>% 
  set_engine("survreg") %>% 
  match(Surv(check_time, standing) ~ relies upon + imports + doc_size + r_size + 
        ns_import + ns_export, 
      knowledge = check_time_train)
survreg_fit(sr_fit)

# A tibble: 7 x 7
  time period         estimate std.error statistic  p.worth conf.low conf.excessive
  <chr>           <dbl>     <dbl>     <dbl>    <dbl>    <dbl>     <dbl>
1 (Intercept)  4.05      0.0174     234.    0.             NA        NA
2 relies upon      0.108     0.00701     15.4   3.40e-53       NA        NA
3 imports      0.0660    0.00327     20.2   1.09e-90       NA        NA
4 doc_size     7.76      0.543       14.3   2.24e-46       NA        NA
5 r_size       0.812     0.0889       9.13  6.94e-20       NA        NA
6 ns_import    0.00501   0.00103      4.85  1.22e- 6       NA        NA
7 ns_export   -0.000212  0.000375    -0.566 5.71e- 1       NA        NA

survreg_pred <- 
  predict(survreg_fit, check_time_test) %>% 
  bind_cols(check_time_test %>% choose(check_time, standing))  

ggplot(survreg_pred, aes(x = check_time, y = .pred, coloration = issue(standing))) +
  geom_point() + 
  coord_cartesian(ylim = c(0, 1400))

Determine 3: Take a look at set predictions from surv_reg. One outlier (of worth 160421) is excluded by way of coord_cartesian() to keep away from distorting the plot.

For the MCMC mannequin, we re-train on simply the coaching set and acquire the parameter abstract. The code is analogous to the above and never proven right here.

We are able to now predict on the take a look at set, for simplicity simply utilizing the posterior means:

df <- check_time_test %>% choose(
                    relies upon,
                    imports,
                    doc_size,
                    r_size,
                    ns_import,
                    ns_export) %>%
  add_column(intercept = rep(1, nrow(check_time_test)), .earlier than = 1)

mcmc_pred <- df %>% as.matrix() %*% abstract$imply %>% exp() %>% as.numeric()
mcmc_pred <- check_time_test %>% choose(check_time, standing) %>%
  add_column(.pred = mcmc_pred)

ggplot(mcmc_pred, aes(x = check_time, y = .pred, coloration = issue(standing))) +
  geom_point() + 
  coord_cartesian(ylim = c(0, 1400)) 

Determine 4: Take a look at set predictions from the mcmc mannequin. No outliers, simply utilizing similar scale as above for comparability.

This seems to be good!

Wrapup

We’ve proven easy methods to mannequin censored knowledge – or fairly, a frequent subtype thereof involving durations – utilizing tfprobability. The check_times knowledge from parsnip have been a enjoyable selection, however this modeling method could also be much more helpful when censoring is extra substantial. Hopefully his submit has offered some steering on easy methods to deal with censored knowledge in your personal work. Thanks for studying!