Datasets:
scenario_id stringlengths 6 6 | scarcity stringclasses 3
values | severity stringclasses 2
values | policy stringclasses 9
values | n_patients int64 16 62 | fulfilment float64 0.14 1 | unmet_transport float64 0 1 | unmet_independent float64 0 1 | cohort_gap float64 0 1 | mean_burden float64 9.88 55 | time_changes float64 0 0.63 | transport_overbooked int64 0 9 | chair_overbooked int64 0 0 | travel_violations int64 0 0 | feasible float64 0 1 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
S01000 | low | partial | list_order_greedy | 26 | 1 | 0 | 0 | 0 | 26.076923 | 0.192308 | 0 | 0 | 0 | 1 |
S01000 | low | partial | most_constrained_greedy | 26 | 1 | 0 | 0 | 0 | 25.396154 | 0.115385 | 0 | 0 | 0 | 1 |
S01000 | low | partial | independent_units | 26 | 0.961538 | 0 | 0.066667 | 0.066667 | 25.128 | 0.192308 | 1 | 0 | 0 | 0 |
S01000 | low | partial | independent_units_gated | 26 | 0.923077 | 0.090909 | 0.066667 | 0.024242 | 25.2375 | 0.192308 | 0 | 0 | 0 | 1 |
S01000 | low | partial | quota_units | 26 | 0.884615 | 0.272727 | 0 | 0.272727 | 26.934783 | 0.192308 | 0 | 0 | 0 | 1 |
S01000 | low | partial | repeated_independent_units | 26 | 1 | 0 | 0 | 0 | 25.892308 | 0.192308 | 1 | 0 | 0 | 0 |
S01000 | low | partial | coordinated_units | 26 | 1 | 0 | 0 | 0 | 26.65 | 0.230769 | 0 | 0 | 0 | 1 |
S01000 | low | partial | optimal_throughput | 26 | 1 | 0 | 0 | 0 | 23.526923 | 0.038462 | 0 | 0 | 0 | 1 |
S01000 | low | partial | optimal_fair | 26 | 1 | 0 | 0 | 0 | 23.526923 | 0.038462 | 0 | 0 | 0 | 1 |
S01001 | low | partial | list_order_greedy | 28 | 1 | 0 | 0 | 0 | 21.971429 | 0.071429 | 0 | 0 | 0 | 1 |
S01001 | low | partial | most_constrained_greedy | 28 | 1 | 0 | 0 | 0 | 21.971429 | 0.071429 | 0 | 0 | 0 | 1 |
S01001 | low | partial | independent_units | 28 | 1 | 0 | 0 | 0 | 20.725 | 0.107143 | 0 | 0 | 0 | 1 |
S01001 | low | partial | independent_units_gated | 28 | 1 | 0 | 0 | 0 | 20.725 | 0.107143 | 0 | 0 | 0 | 1 |
S01001 | low | partial | quota_units | 28 | 1 | 0 | 0 | 0 | 21.553571 | 0.107143 | 0 | 0 | 0 | 1 |
S01001 | low | partial | repeated_independent_units | 28 | 1 | 0 | 0 | 0 | 20.725 | 0.107143 | 0 | 0 | 0 | 1 |
S01001 | low | partial | coordinated_units | 28 | 1 | 0 | 0 | 0 | 20.725 | 0.107143 | 0 | 0 | 0 | 1 |
S01001 | low | partial | optimal_throughput | 28 | 1 | 0 | 0 | 0 | 20.053571 | 0.035714 | 0 | 0 | 0 | 1 |
S01001 | low | partial | optimal_fair | 28 | 1 | 0 | 0 | 0 | 20.053571 | 0.035714 | 0 | 0 | 0 | 1 |
S01002 | low | partial | list_order_greedy | 18 | 0.944444 | 0.166667 | 0 | 0.166667 | 21.558824 | 0.111111 | 0 | 0 | 0 | 1 |
S01002 | low | partial | most_constrained_greedy | 18 | 0.944444 | 0.166667 | 0 | 0.166667 | 22.2 | 0.055556 | 0 | 0 | 0 | 1 |
S01002 | low | partial | independent_units | 18 | 1 | 0 | 0 | 0 | 21.311111 | 0 | 2 | 0 | 0 | 0 |
S01002 | low | partial | independent_units_gated | 18 | 0.888889 | 0.333333 | 0 | 0.333333 | 22.10625 | 0 | 0 | 0 | 0 | 1 |
S01002 | low | partial | quota_units | 18 | 0.777778 | 0.666667 | 0 | 0.666667 | 19.028571 | 0 | 0 | 0 | 0 | 1 |
S01002 | low | partial | repeated_independent_units | 18 | 1 | 0 | 0 | 0 | 21.311111 | 0 | 2 | 0 | 0 | 0 |
S01002 | low | partial | coordinated_units | 18 | 0.944444 | 0.166667 | 0 | 0.166667 | 22.441176 | 0.055556 | 0 | 0 | 0 | 1 |
S01002 | low | partial | optimal_throughput | 18 | 0.944444 | 0.166667 | 0 | 0.166667 | 20.764706 | 0.055556 | 0 | 0 | 0 | 1 |
S01002 | low | partial | optimal_fair | 18 | 0.944444 | 0.166667 | 0 | 0.166667 | 20.764706 | 0.055556 | 0 | 0 | 0 | 1 |
S01003 | low | partial | list_order_greedy | 21 | 1 | 0 | 0 | 0 | 23.447619 | 0.095238 | 0 | 0 | 0 | 1 |
S01003 | low | partial | most_constrained_greedy | 21 | 1 | 0 | 0 | 0 | 23.290476 | 0.095238 | 0 | 0 | 0 | 1 |
S01003 | low | partial | independent_units | 21 | 1 | 0 | 0 | 0 | 21.604762 | 0.047619 | 2 | 0 | 0 | 0 |
S01003 | low | partial | independent_units_gated | 21 | 0.904762 | 0.222222 | 0 | 0.222222 | 21.757895 | 0.047619 | 0 | 0 | 0 | 1 |
S01003 | low | partial | quota_units | 21 | 0.952381 | 0.111111 | 0 | 0.111111 | 23.345 | 0.095238 | 0 | 0 | 0 | 1 |
S01003 | low | partial | repeated_independent_units | 21 | 1 | 0 | 0 | 0 | 21.604762 | 0.047619 | 2 | 0 | 0 | 0 |
S01003 | low | partial | coordinated_units | 21 | 1 | 0 | 0 | 0 | 22.952381 | 0.142857 | 0 | 0 | 0 | 1 |
S01003 | low | partial | optimal_throughput | 21 | 1 | 0 | 0 | 0 | 22.1 | 0.095238 | 0 | 0 | 0 | 1 |
S01003 | low | partial | optimal_fair | 21 | 1 | 0 | 0 | 0 | 22.1 | 0.095238 | 0 | 0 | 0 | 1 |
S01004 | low | partial | list_order_greedy | 25 | 1 | 0 | 0 | 0 | 25.928 | 0.16 | 0 | 0 | 0 | 1 |
S01004 | low | partial | most_constrained_greedy | 25 | 1 | 0 | 0 | 0 | 25.62 | 0.16 | 0 | 0 | 0 | 1 |
S01004 | low | partial | independent_units | 25 | 1 | 0 | 0 | 0 | 24.04 | 0.04 | 2 | 0 | 0 | 0 |
S01004 | low | partial | independent_units_gated | 25 | 0.92 | 0.25 | 0 | 0.25 | 23.786957 | 0.04 | 0 | 0 | 0 | 1 |
S01004 | low | partial | quota_units | 25 | 0.92 | 0.25 | 0 | 0.25 | 24.2 | 0.08 | 0 | 0 | 0 | 1 |
S01004 | low | partial | repeated_independent_units | 25 | 1 | 0 | 0 | 0 | 24.04 | 0.04 | 2 | 0 | 0 | 0 |
S01004 | low | partial | coordinated_units | 25 | 1 | 0 | 0 | 0 | 25.256 | 0.12 | 0 | 0 | 0 | 1 |
S01004 | low | partial | optimal_throughput | 25 | 1 | 0 | 0 | 0 | 24.216 | 0.08 | 0 | 0 | 0 | 1 |
S01004 | low | partial | optimal_fair | 25 | 1 | 0 | 0 | 0 | 24.216 | 0.08 | 0 | 0 | 0 | 1 |
S01005 | low | partial | list_order_greedy | 25 | 1 | 0 | 0 | 0 | 23.628 | 0.12 | 0 | 0 | 0 | 1 |
S01005 | low | partial | most_constrained_greedy | 25 | 1 | 0 | 0 | 0 | 23.936 | 0.12 | 0 | 0 | 0 | 1 |
S01005 | low | partial | independent_units | 25 | 1 | 0 | 0 | 0 | 24.32 | 0.12 | 2 | 0 | 0 | 0 |
S01005 | low | partial | independent_units_gated | 25 | 0.92 | 0.25 | 0 | 0.25 | 24.147826 | 0.12 | 0 | 0 | 0 | 1 |
S01005 | low | partial | quota_units | 25 | 0.96 | 0.125 | 0 | 0.125 | 24.758333 | 0.16 | 0 | 0 | 0 | 1 |
S01005 | low | partial | repeated_independent_units | 25 | 1 | 0 | 0 | 0 | 24.32 | 0.12 | 2 | 0 | 0 | 0 |
S01005 | low | partial | coordinated_units | 25 | 1 | 0 | 0 | 0 | 25.888 | 0.2 | 0 | 0 | 0 | 1 |
S01005 | low | partial | optimal_throughput | 25 | 1 | 0 | 0 | 0 | 22.52 | 0.08 | 0 | 0 | 0 | 1 |
S01005 | low | partial | optimal_fair | 25 | 1 | 0 | 0 | 0 | 22.52 | 0.08 | 0 | 0 | 0 | 1 |
S01006 | low | partial | list_order_greedy | 26 | 1 | 0 | 0 | 0 | 28.884615 | 0.230769 | 0 | 0 | 0 | 1 |
S01006 | low | partial | most_constrained_greedy | 26 | 1 | 0 | 0 | 0 | 28.884615 | 0.230769 | 0 | 0 | 0 | 1 |
S01006 | low | partial | independent_units | 26 | 1 | 0 | 0 | 0 | 28.476923 | 0.153846 | 3 | 0 | 0 | 0 |
S01006 | low | partial | independent_units_gated | 26 | 0.884615 | 0.272727 | 0 | 0.272727 | 29.469565 | 0.153846 | 0 | 0 | 0 | 1 |
S01006 | low | partial | quota_units | 26 | 0.923077 | 0.181818 | 0 | 0.181818 | 28.879167 | 0.115385 | 0 | 0 | 0 | 1 |
S01006 | low | partial | repeated_independent_units | 26 | 1 | 0 | 0 | 0 | 28.476923 | 0.153846 | 3 | 0 | 0 | 0 |
S01006 | low | partial | coordinated_units | 26 | 1 | 0 | 0 | 0 | 30.430769 | 0.269231 | 0 | 0 | 0 | 1 |
S01006 | low | partial | optimal_throughput | 26 | 1 | 0 | 0 | 0 | 27.730769 | 0.192308 | 0 | 0 | 0 | 1 |
S01006 | low | partial | optimal_fair | 26 | 1 | 0 | 0 | 0 | 27.730769 | 0.192308 | 0 | 0 | 0 | 1 |
S01007 | low | partial | list_order_greedy | 17 | 1 | 0 | 0 | 0 | 35.082353 | 0.058824 | 0 | 0 | 0 | 1 |
S01007 | low | partial | most_constrained_greedy | 17 | 1 | 0 | 0 | 0 | 35.1 | 0.058824 | 0 | 0 | 0 | 1 |
S01007 | low | partial | independent_units | 17 | 1 | 0 | 0 | 0 | 37.988235 | 0.235294 | 0 | 0 | 0 | 1 |
S01007 | low | partial | independent_units_gated | 17 | 1 | 0 | 0 | 0 | 37.988235 | 0.235294 | 0 | 0 | 0 | 1 |
S01007 | low | partial | quota_units | 17 | 0.941176 | 0.2 | 0 | 0.2 | 37.50625 | 0.235294 | 0 | 0 | 0 | 1 |
S01007 | low | partial | repeated_independent_units | 17 | 1 | 0 | 0 | 0 | 37.988235 | 0.235294 | 0 | 0 | 0 | 1 |
S01007 | low | partial | coordinated_units | 17 | 1 | 0 | 0 | 0 | 37.988235 | 0.235294 | 0 | 0 | 0 | 1 |
S01007 | low | partial | optimal_throughput | 17 | 1 | 0 | 0 | 0 | 33.9 | 0.058824 | 0 | 0 | 0 | 1 |
S01007 | low | partial | optimal_fair | 17 | 1 | 0 | 0 | 0 | 33.9 | 0.058824 | 0 | 0 | 0 | 1 |
S01008 | low | partial | list_order_greedy | 19 | 0.947368 | 0.2 | 0 | 0.2 | 26.311111 | 0.105263 | 0 | 0 | 0 | 1 |
S01008 | low | partial | most_constrained_greedy | 19 | 0.947368 | 0.2 | 0 | 0.2 | 26.311111 | 0.105263 | 0 | 0 | 0 | 1 |
S01008 | low | partial | independent_units | 19 | 1 | 0 | 0 | 0 | 26.184211 | 0.105263 | 2 | 0 | 0 | 0 |
S01008 | low | partial | independent_units_gated | 19 | 0.894737 | 0.4 | 0 | 0.4 | 26.241176 | 0.105263 | 0 | 0 | 0 | 1 |
S01008 | low | partial | quota_units | 19 | 0.894737 | 0.4 | 0 | 0.4 | 25.694118 | 0.105263 | 0 | 0 | 0 | 1 |
S01008 | low | partial | repeated_independent_units | 19 | 1 | 0 | 0 | 0 | 26.184211 | 0.105263 | 2 | 0 | 0 | 0 |
S01008 | low | partial | coordinated_units | 19 | 0.947368 | 0.2 | 0 | 0.2 | 27.316667 | 0.157895 | 0 | 0 | 0 | 1 |
S01008 | low | partial | optimal_throughput | 19 | 0.947368 | 0.2 | 0 | 0.2 | 25.738889 | 0.105263 | 0 | 0 | 0 | 1 |
S01008 | low | partial | optimal_fair | 19 | 0.947368 | 0.2 | 0 | 0.2 | 25.738889 | 0.105263 | 0 | 0 | 0 | 1 |
S01009 | low | partial | list_order_greedy | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 26.781818 | 0.130435 | 0 | 0 | 0 | 1 |
S01009 | low | partial | most_constrained_greedy | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 26.045455 | 0.130435 | 0 | 0 | 0 | 1 |
S01009 | low | partial | independent_units | 23 | 0.913043 | 0.076923 | 0.1 | 0.023077 | 27.814286 | 0.304348 | 2 | 0 | 0 | 0 |
S01009 | low | partial | independent_units_gated | 23 | 0.826087 | 0.230769 | 0.1 | 0.130769 | 28.868421 | 0.304348 | 0 | 0 | 0 | 1 |
S01009 | low | partial | quota_units | 23 | 0.869565 | 0.230769 | 0 | 0.230769 | 27.81 | 0.304348 | 0 | 0 | 0 | 1 |
S01009 | low | partial | repeated_independent_units | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 28.722727 | 0.304348 | 2 | 0 | 0 | 0 |
S01009 | low | partial | coordinated_units | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 30.036364 | 0.391304 | 0 | 0 | 0 | 1 |
S01009 | low | partial | optimal_throughput | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 24.745455 | 0.086957 | 0 | 0 | 0 | 1 |
S01009 | low | partial | optimal_fair | 23 | 0.956522 | 0.076923 | 0 | 0.076923 | 24.745455 | 0.086957 | 0 | 0 | 0 | 1 |
S01010 | low | partial | list_order_greedy | 28 | 1 | 0 | 0 | 0 | 25.639286 | 0.142857 | 0 | 0 | 0 | 1 |
S01010 | low | partial | most_constrained_greedy | 28 | 1 | 0 | 0 | 0 | 25.753571 | 0.142857 | 0 | 0 | 0 | 1 |
S01010 | low | partial | independent_units | 28 | 1 | 0 | 0 | 0 | 25.842857 | 0.178571 | 1 | 0 | 0 | 0 |
S01010 | low | partial | independent_units_gated | 28 | 0.964286 | 0.071429 | 0 | 0.071429 | 25.874074 | 0.178571 | 0 | 0 | 0 | 1 |
S01010 | low | partial | quota_units | 28 | 0.857143 | 0.285714 | 0 | 0.285714 | 22.704167 | 0.071429 | 0 | 0 | 0 | 1 |
S01010 | low | partial | repeated_independent_units | 28 | 1 | 0 | 0 | 0 | 25.842857 | 0.178571 | 1 | 0 | 0 | 0 |
S01010 | low | partial | coordinated_units | 28 | 1 | 0 | 0 | 0 | 27.114286 | 0.214286 | 0 | 0 | 0 | 1 |
S01010 | low | partial | optimal_throughput | 28 | 1 | 0 | 0 | 0 | 25.096429 | 0.142857 | 0 | 0 | 0 | 1 |
S01010 | low | partial | optimal_fair | 28 | 1 | 0 | 0 | 0 | 25.096429 | 0.142857 | 0 | 0 | 0 | 1 |
S01011 | low | partial | list_order_greedy | 19 | 1 | 0 | 0 | 0 | 25.710526 | 0.157895 | 0 | 0 | 0 | 1 |
DialysisFlow-AgentBench
1,200 synthetic dialysis networks in which one unit closes some or all of its shifts and its patients must be re-placed in free chairs at nearby units, with a shared patient-transport fleet. The dataset also holds the result of nine placement policies on every network, including two exact MILP optima.
Synthetic operational data only. No patient data, no clinical information. Nothing here supports clinical or treatment decisions.
- Code, paper and full method: github.com/MHKhan50/dialysisflow-agentbench
- Software DOI (all versions): 10.5281/zenodo.24354744
- Matches release v0.2.1. Every scenario is regenerated from its seed by
scripts/export_hf.py.
Configurations
scenarios (1,200 rows, 41,794 displaced patients)
| Field | Meaning |
|---|---|
scenario_id |
S01000 to S02199; the number is the generator seed |
scarcity |
low, medium or high: 60–75%, 75–85% or 85–95% of chairs already in use |
severity |
partial (two shifts closed) or full (all four) |
units |
receiving units: unit, position x_km, y_km (the closed unit is at 20, 20) and free_chairs per shift |
transport_seats |
spare seats in the shared transport fleet, per shift, for the whole network |
patients |
displaced patients: pid, current shift, cohort (transport or independent), travel_limit_min (45 or 75), travel_min to each unit (a list in the same order as units), and options, the feasible `unit |
Shifts are MWF-AM, MWF-PM, TTS-AM, TTS-PM. A patient keeps their weekday pattern; moving between morning and afternoon is allowed and adds 20 minutes of burden. Travel time is 5 minutes plus 1.5 minutes per km.
policy_results (10,800 rows: 1,200 scenarios × 9 policies)
scenario_id, scarcity, severity, policy, n_patients, fulfilment (share placed), unmet_transport, unmet_independent, cohort_gap, mean_burden (minutes), time_changes, transport_overbooked, chair_overbooked, travel_violations, feasible (1 if the plan respects every chair, seat and travel limit).
Policies: list_order_greedy, most_constrained_greedy, independent_units, independent_units_gated, quota_units, repeated_independent_units, coordinated_units, optimal_throughput, optimal_fair.
Use
from datasets import load_dataset
scenarios = load_dataset("MHkhan50/dialysisflow-agentbench", "scenarios", split="train")
results = load_dataset("MHkhan50/dialysisflow-agentbench", "policy_results", split="train")
A new placement policy can be scored against the same networks and compared with the exact optima in policy_results.
Limitations
The networks are stylised: straight-line travel, one closure at a time, a fixed transport share of about 35%, and no clinical priority between patients. Results describe these networks, not any real dialysis service.
Citation
@software{hussain_dialysisflow_agentbench,
author = {Hussain, Muzammal},
title = {DialysisFlow-AgentBench: re-placing displaced dialysis patients after a unit closure},
version = {0.2.1},
doi = {10.5281/zenodo.24354744},
url = {https://github.com/MHKhan50/dialysisflow-agentbench}
}
- Downloads last month
- -