Sylvics Apply for the beta
Port Authority open vessel-delay dataReference
REFERENCE

Port Authority open vessel-delay data

๐Ÿ“Œ TL;DR / Quick Summary <div style="background: rgba(0,0,0,0.3); border: 1px dashed rgba(255,255,255,0.2); padding: 15px; border-radius: 8px; color: rgba(255,255,255,0.7); margin-bottom: 30px;"> The open dataset behind ๐Ÿ“Œ Statistics II - Project - What predicts vessel delay at Greywater: fourteen months of vessel calls published by ๐Ÿข Port Vellan Port Authority. Fictional, like the port. One row per call. Fields, units, and the two columns that mean nearly the same thing and are not the same thing. </div>

๐Ÿ› ๏ธ Instructions / Core Material

Column Unit Meaning Trap
call_id none One vessel call A few appear twice, once per berth
vessel_type category Container, bulk, general, tanker Gen. cargo and general are one type
length_m metres Vessel length Clean
gross_tonnage tonnes Size Moves with length; do not use both
scheduled_arrival date and time Planned arrival Three date formats
actual_arrival date and time Arrival at the pilot station Same three
berth_time date and time Alongside Sometimes before arrival, which is impossible
arrival_delay_h hours actual_arrival minus scheduled_arrival Negative means early
berth_delay_h hours Wait between arrival and berth Not the same as the one above
terminal category Which terminal Greywater is about half the rows
wind_class 0 to 4 Wind on the day Recorded once a day, at 06:00

The three date formats are day first, month first, and one summer stretch where somebody typed the month as a word. Port delay analysis script opens with a function that does nothing but admit this.

โš ๏ธ Key Warnings / Constraints

  • arrival_delay_h is the ship being late. berth_delay_h is the port being full. They move together most weeks and mean different things. The project question is about the first one, and I had to say so on slide two.
  • Rows where the vessel berths before it arrives are dropped, counted, and the count is in the appendix.
  • Fourteen months, one port. Whatever it shows, it shows for here.
  • The file is open and has no people in it. It is still a colleague of Hana Kobayashi's who maintains it, so the report is polite about the dates.

๐Ÿ”— Related Entities & Resources

  • ๐Ÿ“Œ Statistics II - Project - What predicts vessel delay at Greywater: the course project that uses this file
  • ๐Ÿข Port Vellan Port Authority: the publisher of the dataset
  • Port delay analysis script: the notebook that cleans the dates and fits the model