Skip to contents

Track Contract Trajectories

Source: R/trajectory_analysis.R
track_contract_trajectories.Rd

Tracks employment trajectories for individuals who experienced a specific contract type, following their employment status for 4 quarters (91 days each) after the end date of their chronologically first occurrence of that contract type.

Usage

track_contract_trajectories(
  data,
  contract_type_var = "COD_TIPOLOGIA_CONTRATTUALE",
  contract_type_value,
  n_quarters = 4,
  person_id = "cf",
  start_date_col = "inizio",
  end_date_col = "fine",
  arco_col = "arco",
  employment_thresholds = c(1, 40, 99),
  return_plot = TRUE,
  palette = "employment",
  use_bw = FALSE,
  plot_title = NULL,
  plot_subtitle = NULL,
  person_vars = NULL,
  chunk_size = 10000,
  language = "it"
)

Arguments

data

A data.table containing vecshift-processed employment data

contract_type_var

Character. Column name containing contract type codes (default: "COD_TIPOLOGIA_CONTRATTUALE")

contract_type_value

Character. Specific contract type value to track

n_quarters

Integer. Number of quarters to track after reference date (default: 4)

person_id

Character. Column name for person identifier (default: "cf")

start_date_col

Character. Column name for contract start dates (default: "inizio")

end_date_col

Character. Column name for contract end dates (default: "fine")

arco_col

Character. Column name for employment status indicator where 0 = unemployment and > 0 = employment (default: "arco")

employment_thresholds

Numeric vector. Three percentage values defining employment status classification thresholds (default: c(1, 40, 99))

return_plot

Logical. Whether to return the alluvial plot (default: TRUE)

palette

Character. Color palette to use: "employment", "main", "desaturated", "bw" (default: "employment")

use_bw

Logical. Force black and white palette (default: FALSE)

plot_title

Character. Custom plot title (default: auto-generated)

plot_subtitle

Character. Custom plot subtitle (default: auto-generated)

person_vars

Character vector of person-level variable names to include in the output data (e.g., c("eta", "sesso", "istruzione")). Variables are extracted from the first target contract for each person. Default is NULL (no additional variables).

chunk_size

Integer. Number of persons to process per chunk for memory management (default: 10000)

language

Character. Language for status labels: "it" for Italian, "en" for English (default: "it")

Value

A list containing:

  • data: data.table with individual-level quarterly trajectories

  • summary: data.table with aggregated trajectory patterns

  • transitions: data.table with aggregated transition counts between quarters, including columns: from_quarter, to_quarter, from_status, to_status, person_vars (if provided), count

  • plot: ggplot2 alluvial plot (if return_plot = TRUE)

  • parameters: List of function parameters used

Details

This function identifies individuals who had a specific contract type, uses the end date of their chronologically FIRST occurrence (earliest start date) as a reference point, then tracks employment status in the subsequent quarters. Employment status is calculated based on the percentage of days employed within each quarter, with overlapping contracts counting as employment time.

Employment is determined by the arco variable where:

  • arco = 0: Indicates unemployment (not working)

  • arco > 0: Indicates employment (working)

All contract periods (both arco = 0 and arco > 0) are included in the analysis, but only periods where arco > 0 are counted as employment days. Employment status is then classified based on the percentage of days employed within each 91-day quarter using configurable thresholds (default: 1%, 40%, 99%):

  • "Not Working": < employment_thresholds1% of days employed

  • "Partially Working": >= employment_thresholds1% and <= employment_thresholds2% of days employed

  • "Mostly Working": > employment_thresholds2% and <= employment_thresholds3% of days employed

  • "Fully Working": > employment_thresholds3% of days employed

  • "No Information": No contract data available for the quarter (days_employed = NA)

The function only tracks forward in time from the reference date. Any employment is counted as working time, regardless of contract intensity or prior employment status. Overlapping contracts are handled by counting the union of employment periods using vectorized operations for maximum performance.

See also

theme_vecshift, vecshift_colors

Examples

if (FALSE) { # \dontrun{
# Track trajectories after temporary contracts
result <- track_contract_trajectories(
  data = employment_data,
  contract_type_value = "B.01.00",
  n_quarters = 4
)

# View the summary of trajectory patterns
print(result$summary)

# Display the alluvial plot
print(result$plot)

# Custom visualization with black and white
result_bw <- track_contract_trajectories(
  data = employment_data,
  contract_type_value = "B.01.00",
  use_bw = TRUE,
  plot_title = "Employment Trajectories After Temporary Contracts"
)

# Large dataset with chunked processing
result_large <- track_contract_trajectories(
  data = large_employment_data,
  contract_type_value = "B.01.00",
  chunk_size = 5000
)
} # }