Survival Curves with create_survival()

library(vivainsights)
library(dplyr)

Overview

create_survival() produces a Kaplan–Meier survival curve — a non-parametric estimate of how quickly a specific event occurs across a population. In a workforce analytics context typical events include:

First week of after-hours collaboration
First week a collaboration or network metric crosses a threshold
First observed use of a new tool

The function expects one row per person with a pre-computed time (weeks until event or end of observation) and event (1 = event occurred, 0 = censored) column. Use create_survival_prep() to derive these from a Standard Person Query panel dataset.

Reframing “survival” for workforce contexts

In classical survival analysis the event is typically something negative — death, equipment failure — and the y-axis probability represents “still alive / still working”. In workforce analytics the event is usually a positive milestone: first adoption of a tool, first week as a power user, first time a network or collaboration metric crosses a meaningful threshold.

The terminology inverts: “surviving” means not yet having reached the milestone, and the event (“death”) means success — the person converted. It can therefore be more intuitive to read the chart as a time-to-adoption curve, a conversion curve, or a graduation curve:

A curve that drops steeply early → most people reached the milestone quickly.
A curve that stays high → many people had not yet converted by the end of the observation window.
The y-axis → the share of people who have not yet experienced the milestone.

The curve answers: “By week N, what fraction of the population had not yet crossed the threshold?”

Step 1 — Prepare person-level survival data

pq_data is a panel dataset with one row per person per week. create_survival_prep() collapses it to one row per person, recording:

time: the week number at which the event first occurred, or the total number of observed weeks if the event never occurred (censored).
event: 1 if the condition was met in at least one week, 0 otherwise.

Here we define the event as a person’s first week with any after-hours collaboration (After_hours_collaboration_hours > 0):

surv_data <- create_survival_prep(
  data            = pq_data,
  metric          = "After_hours_collaboration_hours",
  event_condition = function(x) x > 0,
  hrvar           = "Organization"
)

glimpse(surv_data)
#> Rows: 300
#> Columns: 4
#> $ PersonId     <chr> "0023c8ce-6939-4188-a9ec-f5c205fc4426", "00a79c9a-639e-42…
#> $ time         <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
#> $ event        <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
#> $ Organization <chr> "Legal", "Finance", "Finance", "Finance", "Research", "Fi…

# Event rate and time distribution
cat("Total persons:  ", nrow(surv_data), "\n")
#> Total persons:   300
cat("Event rate:     ", round(mean(surv_data$event) * 100, 1), "%\n")
#> Event rate:      100 %
cat("Weeks observed: ", range(surv_data$time), "\n")
#> Weeks observed:  1 1

surv_data %>%
  count(event, name = "n") %>%
  mutate(pct = round(100 * n / sum(n), 1),
         label = ifelse(event == 1, "Had event", "Censored"))
#> # A tibble: 1 × 4
#>   event     n   pct label    
#>   <int> <int> <dbl> <chr>    
#> 1     1   300   100 Had event

The persons who never had any after-hours work during the observation window appear as censored (event = 0). The survival curve accounts for this: they contribute information up to their last observed week.

Step 2 — Plot the Kaplan–Meier curve

Pass the person-level data to create_survival(), specifying the time and event columns and the grouping variable:

create_survival(
  data      = surv_data,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 5
)

Reading the chart:

Y-axis — probability of not yet having reached the milestone (the “survival” probability, or equivalently, the share yet to convert).
X-axis — week number (time since the start of observation).
Each step down marks one or more conversions (events) in that group at that week.
Curves that drop quickly indicate groups where most people reached the milestone early (fast adoption).
Curves that stay high indicate groups where many people had not yet converted by the end of the window (slow or incomplete adoption).

Overall curve (no grouping)

Set hrvar = NULL to estimate a single curve across the whole population:

create_survival(
  data      = surv_data,
  time_col  = "time",
  event_col = "event",
  hrvar     = NULL,
  mingroup  = 5
)

Returning the survival table

Set return = "table" to get the underlying long-format data frame. Each row represents one event time within one group:

surv_tbl <- create_survival(
  data      = surv_data,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 5,
  return    = "table"
)

head(surv_tbl, 12)
#>    Organization  n time survival at_risk events
#> 1       Finance 68    0        1      68      0
#> 2       Finance 68    1        0      68     68
#> 3            HR 33    0        1      33      0
#> 4            HR 33    1        0      33     33
#> 5            IT 68    0        1      68      0
#> 6            IT 68    1        0      68     68
#> 7         Legal 44    0        1      44      0
#> 8         Legal 44    1        0      44     44
#> 9    Operations 22    0        1      22      0
#> 10   Operations 22    1        0      22     22
#> 11     Research 52    0        1      52      0
#> 12     Research 52    1        0      52     52

Columns:

Column	Description
`Organization`	Group identifier
`n`	Total persons in group (after privacy filtering)
`time`	Week number
`survival`	Estimated survival probability at that time
`at_risk`	Persons still in the risk set (event not yet occurred)
`events`	Events occurring at this time point

Extracting median event times

The median survival time is the week at which 50 % of the group has experienced the event:

surv_tbl %>%
  group_by(Organization) %>%
  filter(survival <= 0.5) %>%
  slice(1) %>%                         # first time survival crosses 0.5
  ungroup() %>%
  select(Organization, n, time, survival) %>%
  arrange(time)
#> # A tibble: 7 × 4
#>   Organization     n  time survival
#>   <chr>        <int> <dbl>    <dbl>
#> 1 Finance         68     1        0
#> 2 HR              33     1        0
#> 3 IT              68     1        0
#> 4 Legal           44     1        0
#> 5 Operations      22     1        0
#> 6 Research        52     1        0
#> 7 Sales           13     1        0

Groups with a smaller time value here reach the event threshold faster on average.

Grouping by a different HR variable

Any character column can be used as the grouping variable. Here we compare after-hours adoption by LevelDesignation:

surv_level <- create_survival_prep(
  data            = pq_data,
  metric          = "After_hours_collaboration_hours",
  event_condition = function(x) x > 0,
  hrvar           = "LevelDesignation"
)

create_survival(
  data      = surv_level,
  time_col  = "time",
  event_col = "event",
  hrvar     = "LevelDesignation",
  mingroup  = 5
)

Changing the event definition

The event_condition argument in create_survival_prep() accepts any function that returns a logical vector. This makes it easy to explore different thresholds without modifying your data.

Higher after-hours threshold

# Event: first week with more than 2 hours of after-hours collaboration
surv_high <- create_survival_prep(
  data            = pq_data,
  metric          = "After_hours_collaboration_hours",
  event_condition = function(x) x > 2,
  hrvar           = "Organization"
)

cat("Event rate (> 2 h after-hours):", round(mean(surv_high$event) * 100, 1), "%\n")
#> Event rate (> 2 h after-hours): 100 %

create_survival(
  data      = surv_high,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 5
)

Network growth milestone

# Event: first week where internal network size exceeds 10 contacts
surv_net <- create_survival_prep(
  data            = pq_data,
  metric          = "Internal_network_size",
  event_condition = function(x) x > 10,
  hrvar           = "Organization"
)

cat("Event rate (network > 10):", round(mean(surv_net$event) * 100, 1), "%\n")
#> Event rate (network > 10): 100 %

create_survival(
  data      = surv_net,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 5
)

Privacy filtering

Groups below mingroup unique persons are removed before the curve is estimated. Increase the threshold to be more conservative:

surv_strict <- create_survival(
  data      = surv_data,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 20,
  return    = "table"
)

# Which groups remained after stricter filtering?
surv_strict %>%
  distinct(Organization, n) %>%
  arrange(desc(n))
#>   Organization  n
#> 1      Finance 68
#> 2           IT 68
#> 3     Research 52
#> 4        Legal 44
#> 5           HR 33
#> 6   Operations 22

Handling missing values

na.rm = TRUE (the default) drops rows where time_col or event_col are NA before the curve is estimated. Set na.rm = FALSE to keep them; any remaining NA values will be silently excluded during coercion and a warning will be issued.

Using the lower-level helpers directly

Like create_radar(), the survival family exposes its building blocks as exported functions:

create_survival_calc() — accepts person-level data and returns a list with $table (the long survival table) and $counts (group size table).
create_survival_viz() — accepts the $table output and returns a ggplot object.

library(ggplot2)

calc <- create_survival_calc(
  data      = surv_data,
  time_col  = "time",
  event_col = "event",
  hrvar     = "Organization",
  mingroup  = 5
)

# Inspect group sizes
calc$counts
#> # A tibble: 7 × 2
#>   Organization     n
#>   <chr>        <int>
#> 1 Finance         68
#> 2 HR              33
#> 3 IT              68
#> 4 Legal           44
#> 5 Operations      22
#> 6 Research        52
#> 7 Sales           13

# Render with extra annotation
create_survival_viz(
  data     = calc$table,
  hrvar    = "Organization",
  title    = "Time to first after-hours week",
  subtitle = "Kaplan\u2013Meier estimate by organisation"
) +
  geom_hline(yintercept = 0.5, linetype = "dashed", colour = "grey50") +
  annotate("text", x = 0, y = 0.52, label = "50 % threshold",
           hjust = 0, size = 3, colour = "grey50")

The dashed line at 0.5 makes it easy to read off the median time-to-event for each group.

vivainsights

2026-03-23