# ===================================================================== # Mobility in Germany 2023 (MiD 2023) analysis: Driving Licences, Mileage and # Travel Behaviour of the Population Aged 65+ # # Replication code for: https://doi.org/10.1016/j.tranpol.2026.104493 # # Data source: MiD 2023, Standard data package (B1 v1.1) # infas, DLR, IVT & infas 360 (2025) on behalf of the # German Federal Ministry for Transport (BMV). # Obtained from https://mobilithek.info/offers/897527161045086208. # Not redistributable. # # Required input files (place in the working directory): # - MiD2023_Personen.csv (person-level file) # - MiD2023_Wege.csv (trip-level file) # # Outputs: # - result_modes_sm.xlsx # - result_modes_sm_geo.xlsx # - MiD_65plus_FS_RegioStaR2_DistClass_Summary.xlsx # # Tested with R 4.6.1, tidyverse 2.0.0, writexl 2.0.1 # ===================================================================== rm(list = ls()) library(tidyverse) library(writexl) #--- Set own working directory ---------------------------------------- setwd(".....") # ===================================================================== # 1. CONFIGURATION # ===================================================================== # --- Paths ----------------------------------------------------------- path_persons <- "MiD2023_Personen.csv" path_trips <- "MiD2023_Wege.csv" path_output <- "." # --- Target ages for the age-specific licence/mileage profile -------- # Ages 65-80 are available exactly; HP_ALTER == 85 pools everyone 85+, # so 85, 90 and 95 all receive the same pooled rate. target_ages <- c(65, 70, 75, 80, 85, 90, 95) ages_exact <- c(65, 70, 75, 80) ages_pooled_85up <- c(85, 90, 95) # --- MiD missing-value codes for P_FS_PKW (car driving licence) ------ # 1 = yes # 2 = no # 9 = no answer -> redistribute proportionally # 206 = proxy interview, adult -> redistribute proportionally # 403 = person under 16 -> redistribute ONLY within 65+ subsets, # where it cannot occur; treated as a # definite "no" for all-age subsets # (see calc_licence_share_allages()). LICENCE_VALID <- c(1, 2) LICENCE_MISSING <- c(9, 206, 403) # --- External population figures, Germany, 2024 (Destatis) ----------- population_2024_by_age <- tibble( age = c(65, 70, 75, 80, 85, 90, 95), population_tot = c( 1169242, # aged 65 949236, # aged 70 797432, # aged 75 607141, # aged 80 582919, # aged 85 233324, # aged 90 58459 # aged 95 ) ) pop_65plus_2024 <- 19005074 pop_17plus_2024 <- 70395447 pop_18plus_2024 <- 69602098 pop_total_2024 <- 83577140 # ===================================================================== # 2. LOAD DATA # ===================================================================== persons <- read.csv(path_persons, sep = ",") trips <- read.csv(path_trips, sep = ",") # ===================================================================== # 3. HELPER FUNCTIONS # ===================================================================== # --------------------------------------------------------------------- # calc_licence_share() # # Share of car-licence holders in a person-level subset. # Among those who gave a valid answer, compute the "yes" share; assume # the same share applies to the missings and add them back in # proportionally. Grossed-up person weights (P_HOCH) are used so that # the share refers to the population, not to the raw sample. # # Use for subsets of adults only (no one under 16). # --------------------------------------------------------------------- calc_licence_share <- function(person_subset) { n_yes <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 1], na.rm = TRUE) n_no <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 2], na.rm = TRUE) n_missing <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW %in% LICENCE_MISSING], na.rm = TRUE) n_valid <- n_yes + n_no share_yes <- n_yes / n_valid licence_share <- (n_yes + share_yes * n_missing) / (n_valid + n_missing) tibble( licence_share = licence_share, n_persons_unweigh = nrow(person_subset), pct_missing = n_missing / (n_valid + n_missing) * 100 ) } # --------------------------------------------------------------------- # calc_licence_share_allages() # # Same logic, but for subsets that include people under 16. For those, # code 403 is not an unknown value but a definite "no" (nobody under 16 # holds a car licence), so it is added to the "no" group instead of being # redistributed. Only codes 9 and 206 remain genuine missings. # --------------------------------------------------------------------- calc_licence_share_allages <- function(person_subset) { n_yes <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 1], na.rm = TRUE) n_no <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 2], na.rm = TRUE) + sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 403], na.rm = TRUE) n_missing <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW %in% c(9, 206)], na.rm = TRUE) n_valid <- n_yes + n_no share_yes <- n_yes / n_valid licence_share <- (n_yes + share_yes * n_missing) / (n_valid + n_missing) tibble( licence_share = licence_share, n_persons_unweigh = nrow(person_subset), pct_missing = n_missing / (n_valid + n_missing) * 100 ) } # --------------------------------------------------------------------- # calc_mileage_rate() # # Average car-driver kilometres per licence holder per day. # # Step 1: driver share among trips where the driver/passenger role is # known; used to scale ambiguous car trips. # Step 2: assign driver kilometres per trip (pkw_fmf): # 1 driver -> full distance # 2 passenger -> 0 # 3 both (driver change) -> half distance # 706 no car used -> 0 # 701/9/708 role unknown but car used (W_VM_G == 1) # -> distance x observed driver share # Step 3: gross up to national daily totals with trip weights (W_HOCH). # Step 4: divide by the grossed-up number of licence holders in the same # subset. Numerator and denominator share the same grossing-up # base, so the resulting rate is independent of population size # and can be applied to any external population figure. # --------------------------------------------------------------------- calc_mileage_rate <- function(trip_subset, person_subset) { # Step 1: observed driver share n_driver <- sum(trip_subset$W_GEW[trip_subset$pkw_fmf == 1], na.rm = TRUE) n_passenger <- sum(trip_subset$W_GEW[trip_subset$pkw_fmf == 2], na.rm = TRUE) driver_share <- n_driver / (n_driver + n_passenger) # Step 2: driver kilometres per trip trip_subset <- trip_subset %>% mutate(car_driver_km = case_when( pkw_fmf == 1 ~ wegkm_imp, pkw_fmf == 2 ~ 0, pkw_fmf == 3 ~ wegkm_imp * 0.5, pkw_fmf == 706 ~ 0, pkw_fmf %in% c(701, 9, 708) & W_VM_G == 1 ~ wegkm_imp * driver_share, TRUE ~ 0 )) # Step 3: national daily car-driver kilometres car_km_day <- sum(trip_subset$car_driver_km * trip_subset$W_HOCH, na.rm = TRUE) # Step 4: grossed-up licence holders in the same subset n_yes <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 1], na.rm = TRUE) n_no <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW == 2], na.rm = TRUE) n_missing <- sum(person_subset$P_HOCH[person_subset$P_FS_PKW %in% LICENCE_MISSING], na.rm = TRUE) n_valid <- n_yes + n_no licence_holders <- n_yes + (n_yes / n_valid) * n_missing tibble( km_per_holder_day = car_km_day / licence_holders, km_per_holder_year = (car_km_day / licence_holders) * 365, n_trips_unweigh = nrow(trip_subset), driver_share = driver_share ) } # ===================================================================== # 4. AGE-SPECIFIC LICENCE SHARES L_i AND MILEAGE RATES M_i # ===================================================================== # --- 4.1 Licence shares by age -------------------------------------- # Exact single ages 65, 70, 75, 80 licence_exact <- map_dfr(ages_exact, function(a) { calc_licence_share(persons %>% filter(HP_ALTER == a)) %>% mutate(age = a, mid_group = as.character(a), .before = 1) }) # Pooled 85+ group, applied to ages 85, 90 and 95 licence_85plus_raw <- calc_licence_share(persons %>% filter(HP_ALTER == 85)) licence_85plus <- tibble(age = ages_pooled_85up) %>% mutate( mid_group = "85+", licence_share = licence_85plus_raw$licence_share, n_persons_unweigh = licence_85plus_raw$n_persons_unweigh, pct_missing = licence_85plus_raw$pct_missing ) licence_shares <- bind_rows(licence_exact, licence_85plus) # --- 4.2 Number of licence holders = share x 2024 population --------- licence_holders_by_age <- population_2024_by_age %>% left_join(licence_shares, by = "age") %>% mutate(licence_holders = population_tot * licence_share) # --- 4.3 Mileage rates by age --------------------------------------- mileage_exact <- map_dfr(ages_exact, function(a) { calc_mileage_rate( trip_subset = trips %>% filter(HP_ALTER == a), person_subset = persons %>% filter(HP_ALTER == a) ) %>% mutate(age = a, mileage_source = "MiD exact", .before = 1) }) mileage_85plus_raw <- calc_mileage_rate( trip_subset = trips %>% filter(HP_ALTER == 85), person_subset = persons %>% filter(HP_ALTER == 85) ) mileage_85plus <- tibble(age = ages_pooled_85up) %>% mutate( mileage_source = "MiD 85+ pooled", km_per_holder_day = mileage_85plus_raw$km_per_holder_day, km_per_holder_year = mileage_85plus_raw$km_per_holder_year, n_trips_unweigh = mileage_85plus_raw$n_trips_unweigh, driver_share = mileage_85plus_raw$driver_share ) mileage_rates <- bind_rows(mileage_exact, mileage_85plus) # --- 4.4 Combined age profile --------------------------------------- age_profile <- licence_holders_by_age %>% left_join(mileage_rates, by = "age") %>% select( age, population_tot, licence_share, licence_holders, mid_group, km_per_holder_day, km_per_holder_year, mileage_source, n_persons_unweigh, pct_missing, n_trips_unweigh, driver_share ) # ===================================================================== # 5. AGGREGATE LICENCE HOLDER COUNTS # ===================================================================== # HP_ALTER >= 65 covers ages 65-84 exactly plus the pooled 85+ category, # i.e. the full 65+ population. # --- 5.1 Total licence holders aged 65+ (L_tot_65plus) -------------- persons_65plus <- persons %>% filter(HP_ALTER >= 65) licence_65plus <- calc_licence_share(persons_65plus) licence_holders_65plus <- pop_65plus_2024 * licence_65plus$licence_share # --- 5.2 Total licence holders aged 17+ (L_tot_17plus) -------------- persons_17plus <- persons %>% filter(HP_ALTER >= 17) licence_17plus <- calc_licence_share(persons_17plus) licence_holders_17plus <- pop_17plus_2024 * licence_17plus$licence_share # --- 5.3 Licence share in the total population ---------------------- # Code 403 (under 16) is treated as a definite "no" here. licence_total_pop <- calc_licence_share_allages(persons) # ===================================================================== # 6. CAR OCCUPANCY, AGED 65+ # ===================================================================== # anzpers = number of people in the vehicle including the driver. # Valid range 1-6 (6 = "6 or more"); 99 = no answer, 701 = not collected # for regular business trips (rbW). Both are excluded. # pkw_fmf == 1 restricts to trips where the respondent was the driver. # W_GEW (not W_HOCH) is used because this is a mean, not a total. occupancy_65plus <- trips %>% filter( HP_ALTER >= 65, pkw_fmf == 1, anzpers %in% 1:6 ) %>% summarise( occupancy = weighted.mean(anzpers, W_GEW), n_trips_unweigh = n() ) # ===================================================================== # 7. TOTAL ANNUAL CAR-DRIVER KILOMETRES # ===================================================================== # The MiD grossing-up factors reproduce the survey-period population # (~2023). To keep these totals consistent with the 2024 population base # used elsewhere, the population-independent MiD rate (km per licence # holder per day) is multiplied by the externally sourced 2024 number of # licence holders rather than being taken directly from W_HOCH sums. # --- 7.1 Aged 65+ (M_tot_65plus) ------------------------------------ trips_65plus <- trips %>% filter(HP_ALTER >= 65) # Observed driver share among 65+ car trips n_driver_65 <- sum(trips_65plus$W_GEW[trips_65plus$pkw_fmf == 1], na.rm = TRUE) n_passenger_65 <- sum(trips_65plus$W_GEW[trips_65plus$pkw_fmf == 2], na.rm = TRUE) driver_share_65 <- n_driver_65 / (n_driver_65 + n_passenger_65) # Driver kilometres per trip (same rule set as calc_mileage_rate) trips_65plus <- trips_65plus %>% mutate(car_driver_km = case_when( pkw_fmf == 1 ~ wegkm_imp, pkw_fmf == 2 ~ 0, pkw_fmf == 3 ~ wegkm_imp * 0.5, pkw_fmf == 706 ~ 0, pkw_fmf %in% c(701, 9, 708) & W_VM_G == 1 ~ wegkm_imp * driver_share_65, TRUE ~ 0 )) # National daily car-driver kilometres on the MiD (~2023) population base car_km_day_65plus <- sum(trips_65plus$car_driver_km * trips_65plus$W_HOCH, na.rm = TRUE) # Licence holders on the same (~2023) base, used as denominator licence_holders_mid_65plus <- licence_65plus$licence_share * sum(persons_65plus$P_HOCH[persons_65plus$P_FS_PKW %in% c(1, 2, 9, 206, 403)], na.rm = TRUE) # Population-independent rate km_per_holder_day_65plus <- car_km_day_65plus / licence_holders_mid_65plus # Rescale to the 2024 population licence_holders_65plus_2024 <- pop_65plus_2024 * licence_65plus$licence_share vkt_year_65plus <- licence_holders_65plus_2024 * km_per_holder_day_65plus * 365 # --- 7.2 Aged 18+ (M_tot_18plus) ------------------------------------ persons_18plus <- persons %>% filter(HP_ALTER >= 18) trips_18plus <- trips %>% filter(HP_ALTER >= 18) licence_18plus <- calc_licence_share(persons_18plus) mileage_18plus <- calc_mileage_rate(trips_18plus, persons_18plus) licence_holders_18plus_2024 <- pop_18plus_2024 * licence_18plus$licence_share vkt_year_18plus <- licence_holders_18plus_2024 * mileage_18plus$km_per_holder_day * 365 # --- 7.3 Average mileage per licence holder, all adults 18+ ---------- mileage_all_drivers <- calc_mileage_rate( trip_subset = trips %>% filter(HP_ALTER >= 18), person_subset = persons %>% filter(HP_ALTER >= 18) ) # ===================================================================== # 8. CARS PER LICENCE HOLDER, AGED 65+ (kappa) # ===================================================================== # Each 65+ licence holder is assigned an equal share of the cars in their # household: household cars / licence holders in that household. The # weighted mean of these shares gives cars per licence holder. # # H_ID == 0 marks persons who cannot be linked to a household and is # excluded. H_ANZAUTO is capped at 3 ("3 or more"); 99 = no answer. # --- 8.1 Licence holders per household ------------------------------ licence_holders_per_hh <- persons %>% filter(H_ID != 0, P_FS_PKW == 1) %>% group_by(H_ID) %>% summarise(n_holders_in_hh = n(), .groups = "drop") # --- 8.2 Car share per 65+ licence holder --------------------------- cars_65plus <- persons %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, H_ID != 0, H_ANZAUTO %in% 0:3 # valid values only ) %>% left_join(licence_holders_per_hh, by = "H_ID") %>% mutate(car_share = H_ANZAUTO / n_holders_in_hh) # --- 8.3 Weighted mean ---------------------------------------------- # P_GEW (not P_HOCH): this is a mean, not a population total. cars_per_licence_holder <- cars_65plus %>% summarise( cars_per_holder = weighted.mean(car_share, P_GEW), n_persons_unweigh = n(), n_unlinked = sum(is.na(n_holders_in_hh)) ) # ===================================================================== # 9. MODE-SPECIFIC TRAVEL BEHAVIOUR, AGED 65+ WITH A LICENCE # ===================================================================== # Modal split, mean trip distance and mean trip duration by main mode. # All three are means/shares, so W_GEW is used throughout. # # hvm_diff1 (main mode, differentiated): # 1 = walk, 2 = bicycle, 5 = car passenger, 6 = car driver, # 8 = local public transport, 9 = taxi, 10 = long-distance public transport # Modes not listed (motorcycle, lorry, other, no answer) are excluded. mode_map <- tibble( hvm_diff1 = c(1, 2, 6, 5, 8, 9, 10), mode = c("Walk", "Bike", "Car Driver", "Car Passenger", "Public Transport", "Taxi", "Long-dist. PT") ) %>% mutate(mode = factor(mode, levels = c( "Walk", "Bike", "Car Driver", "Car Passenger", "Public Transport", "Taxi", "Long-dist. PT" ))) # --- 9.1 Trips of licence holders aged 65+ -------------------------- trips_65_lic <- trips %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, hvm_diff1 %in% mode_map$hvm_diff1 ) %>% left_join(mode_map, by = "hvm_diff1") # --- 9.2 Modal split (share of trips by mode) ----------------------- modal_split <- trips_65_lic %>% group_by(mode) %>% summarise(n_weighted = sum(W_GEW, na.rm = TRUE), .groups = "drop") %>% mutate( share = n_weighted / sum(n_weighted), share_pct = share * 100 ) # --- 9.3 Mean trip distance ----------------------------------------- mean_distance <- trips_65_lic %>% group_by(mode) %>% summarise( mean_km = weighted.mean(wegkm_imp, W_GEW, na.rm = TRUE), .groups = "drop" ) # --- 9.4 Mean trip duration ----------------------------------------- # wegmin_imp1: imputed trip duration in minutes, including rbW trips. mean_duration <- trips_65_lic %>% group_by(mode) %>% summarise( mean_min = weighted.mean(wegmin_imp1, W_GEW, na.rm = TRUE), .groups = "drop" ) # --- 9.5 Combined table --------------------------------------------- modes_65_lic <- modal_split %>% left_join(mean_distance, by = "mode") %>% left_join(mean_duration, by = "mode") %>% select(mode, share_pct, mean_km, mean_min) # Unweighted trip counts for reference n_trips_by_mode <- trips_65_lic %>% count(mode, name = "n_trips_unweigh") write_xlsx(modes_65_lic, file.path(path_output, "result_modes_sm.xlsx")) # ===================================================================== # 10. IMMOBILITY RATE, AGED 65+ WITH A LICENCE # ===================================================================== # Share of persons reporting no trips on their assigned reporting day. # mobil: 0 = not mobile, 1 = mobile, 9 = mobility unknown (excluded). immobility_65_lic <- persons %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, mobil %in% c(0, 1) ) %>% summarise( n_immobile_weighted = sum(P_GEW[mobil == 0]), n_mobile_weighted = sum(P_GEW[mobil == 1]), n_total_weighted = n_immobile_weighted + n_mobile_weighted, immobility_rate = n_immobile_weighted / n_total_weighted, n_persons_unweigh = n() ) # ===================================================================== # 11. RegioStaR2 x DISTANCE CLASS # ===================================================================== # Trips are split at 5 km (<=5 km vs >5 km) within each region type, # giving four groups. The immobility rate is a person-level measure and # therefore cannot be split by distance class; it is reported by region # type only. geo_map_r2 <- tibble( RegioStaR2 = c(1, 2), geo_type = c("Urban", "Rural")) %>% mutate(geo_type = factor(geo_type, levels = c("Urban", "Rural"))) dist_levels <- c("<=5 km", ">5 km") # --- 11.1 Subsets ---------------------------------------------------- persons_65_lic_r2d <- persons %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, RegioStaR2 %in% c(1, 2) ) %>% left_join(geo_map_r2, by = "RegioStaR2") trips_65_lic_r2d <- trips %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, RegioStaR2 %in% c(1, 2), hvm_diff1 %in% mode_map$hvm_diff1 ) %>% left_join(geo_map_r2, by = "RegioStaR2") %>% left_join(mode_map, by = "hvm_diff1") %>% mutate( dist_class = factor(ifelse(wegkm_imp <= 5, "<=5 km", ">5 km"), levels = dist_levels) ) # --- 11.2 Sample sizes ---------------------------------------------- n_persons_r2d <- persons_65_lic_r2d %>% count(geo_type, name = "n_persons_unweighted") n_trips_r2d <- trips_65_lic_r2d %>% count(geo_type, dist_class, name = "n_trips_unweighted") n_trips_r2d_by_mode <- trips_65_lic_r2d %>% count(geo_type, dist_class, mode, name = "n") %>% pivot_wider(names_from = mode, values_from = n, values_fill = 0) # --- 11.3 Modal split ------------------------------------------------ modal_split_r2d <- trips_65_lic_r2d %>% group_by(geo_type, dist_class, mode) %>% summarise(n_weighted = sum(W_GEW, na.rm = TRUE), n_unweigh = n(), .groups = "drop") %>% group_by(geo_type, dist_class) %>% mutate(share_pct = n_weighted / sum(n_weighted) * 100) %>% ungroup() # --- 11.4 Mean distance ---------------------------------------------- mean_distance_r2d <- trips_65_lic_r2d %>% group_by(geo_type, dist_class, mode) %>% summarise( mean_km = weighted.mean(wegkm_imp, W_GEW, na.rm = TRUE), n_unweigh = n(), .groups = "drop" ) # --- 11.5 Mean duration ---------------------------------------------- mean_duration_r2d <- trips_65_lic_r2d %>% group_by(geo_type, dist_class, mode) %>% summarise( mean_min = weighted.mean(wegmin_imp1, W_GEW, na.rm = TRUE), n_unweigh = n(), .groups = "drop" ) # --- 11.6 Immobility rate (region type only) ------------------------- immobility_r2d <- persons_65_lic_r2d %>% filter(mobil %in% c(0, 1)) %>% group_by(geo_type) %>% summarise( n_immobile_weighted = sum(P_GEW[mobil == 0]), n_mobile_weighted = sum(P_GEW[mobil == 1]), immobility_pct = n_immobile_weighted / (n_immobile_weighted + n_mobile_weighted) * 100, n_persons_unweighted = n(), .groups = "drop" ) # --- 11.7 Share of car-driver kilometres ------------------------------ vkt_share_r2d <- trips %>% filter( HP_ALTER >= 65, P_FS_PKW == 1, pkw_fmf == 1, RegioStaR2 %in% c(1, 2) ) %>% left_join(geo_map_r2, by = "RegioStaR2") %>% mutate( dist_class = factor(ifelse(wegkm_imp <= 5, "<=5 km", ">5 km"), levels = dist_levels) ) %>% group_by(geo_type, dist_class) %>% summarise( km_grossed = sum(wegkm_imp * W_HOCH, na.rm = TRUE), n_trips_unweighted = n(), .groups = "drop" ) %>% mutate( km_total = sum(km_grossed), share_of_km_pct = km_grossed / km_total * 100 ) # --- 11.8 Export tables ---------------------------------------------- r2d_combined_long <- modal_split_r2d %>% select(geo_type, dist_class, mode, share_pct) %>% left_join( mean_distance_r2d %>% select(geo_type, dist_class, mode, mean_km), by = c("geo_type", "dist_class", "mode") ) %>% left_join( mean_duration_r2d %>% select(geo_type, dist_class, mode, mean_min, n_unweigh), by = c("geo_type", "dist_class", "mode") ) %>% rename(n_trips_unweighted = n_unweigh) r2d_modal_split_wide <- modal_split_r2d %>% select(geo_type, dist_class, mode, share_pct) %>% pivot_wider(names_from = mode, values_from = share_pct, values_fill = 0) r2d_distance_wide <- mean_distance_r2d %>% select(geo_type, dist_class, mode, mean_km) %>% pivot_wider(names_from = mode, values_from = mean_km, values_fill = NA) r2d_duration_wide <- mean_duration_r2d %>% select(geo_type, dist_class, mode, mean_min) %>% pivot_wider(names_from = mode, values_from = mean_min, values_fill = NA) r2d_immobility <- immobility_r2d %>% select(geo_type, immobility_pct, n_persons_unweighted) r2d_vkt_share <- vkt_share_r2d %>% select(geo_type, dist_class, share_of_km_pct, n_trips_unweighted) r2d_sample_sizes <- n_trips_r2d %>% left_join(n_persons_r2d, by = "geo_type") r2d_sample_detail <- as.data.frame(n_trips_r2d_by_mode) write_xlsx( list( "Combined_Long" = r2d_combined_long, "Modal_Split_Wide" = r2d_modal_split_wide, "Distance_Wide" = r2d_distance_wide, "Duration_Wide" = r2d_duration_wide, "Immobility" = r2d_immobility, "KM_Distribution" = r2d_vkt_share, "Sample_Sizes" = r2d_sample_sizes, "Sample_by_Mode" = r2d_sample_detail ), file.path(path_output, "MiD_65plus_FS_RegioStaR2_DistClass_Summary.xlsx") ) # ===================================================================== # END OF SCRIPT # =====================================================================