Table 1. Temporal performances (time required for computation) for each step of a typical data flow.
| Dataset #1∶105 initial VMS pings | Dataset #2∶106 initial VMS pings | ||||
| General Step of the Analysis | Tool | Time required tocomplete the processing(minutes) | Main Statistics | Time required tocomplete theprocessing (minutes) | Main Statistics |
| VMS Data Management | Edit RawData | <1 | 0 NAs found in latitude degrees;0 NAs found in latitude minutes;0 NAs found in latitude seconds;0 NAs found in latitude direction;0 Latitudes out of range (−90/90);0 NAs found in longitude degrees;0 NAs found in longitude minutes;0 NAs found in longitude seconds;0 NAs found in longitude direction;0 Longitudes out of range (−180/180);0 NAs found in dates; 0 NAs found inhours; 0 NAs found in minutes; 0 NAsfound in seconds; 0 dates found withbad format; 0 NAs found in knotsspeed; 0 NAs found in degrees heading | <1 | 0 NAs found in latitude degrees; 0 NAs found inlatitude minutes; 0 NAs found in latitude seconds;0 NAs found in latitude direction; 0 Latitudes outof range (−90/90); 0 NAs found in longitude degrees;0 NAs found in longitude minutes; 0 NAs found inlongitude seconds; 0 NAs found in longitude direction;0 Longitudes out of range (−180/180); 0 NAs found indates; 0 NAs found in hours; 0 NAs found in minutes;0 NAs found in seconds; 0 dates found with badformat; 0 NAs found in knots speed; 0 NAs found in degrees heading |
| CreateDatabase | <1 | / | <1 | / | |
| Load DBin theVMS DataViewer | <1 | / | <1 | / | |
| Clean DBData | 20 | Found 4584 (4.58% of total) duplicatedpings; Found 31019 (31.02% of total)pings in harbour; Found 7550 (7.55%of total) pings on land; Found 38(0.04% of total) not coherent pings | 150 | Found 98381 (9.84% of total) duplicated pings;Found 272996 (27.31% of total) pings in harbour;Found 124015 (12.41% of total) pings on land;Found 279 (0.03% of total) not coherent pings | |
| Track Cutting | 5 | 3883 tracks detected | 40 | 36994 tracks detected | |
| Interpolation(10 minutesfrequency) | 20 | Number of pings changed increasesfrom 105 (real) to 4.9*105 (interpolated) | 180 | Number of pings changed increases from 106(real) to 4.4*106 (interpolated) | |
| Assign Bathymetry(1 degree resolution) | 150 (Fast& Heavy Algorithm)/60(Slow & Light) | / | 1000 (Fast& HeavyAlgorithm)/400(Slow & Light)$ | / | |
| Assign Area(MediterraneanGSAs) | 100 | / | 900 | / | |
| Subtotal | 295 (∼5 hours) | 2237 | 2270 (∼37 hours) | ||
| Logbook Data Management (2×105 initial records) | Tool | Time required to complete the processing (minutes) | Main Statistics | ||
| Edit Raw Data | <1 | 20 NAs found in Start Times; 140 NAs found in End Times;0 NAs found in Species; 0 NAs found in Quantity;Removed 0.08% of data, that is 160 logbooks | |||
| Create Database | 180 | This step requires only few minutes for the EFLALO format | |||
| Métier Discovery(searching between2–30 groups on thewhole dataset with100 samples of 1000records each) | 20 | The best partitioning corresponded to 11 métiers. These are alsoavailable into to the package as reference dataset for Métier Classification | |||
| Métier editing | Depends by the user, reasonablyno more than half an hour | 69397 records in the database; 387 species | |||
| Métier Classification | 20 | / | |||
| Subtotal | ∼250 (∼ 4 hours) | ||||
| VMS-Logbook Analysis | Tool | Time required to complete the processing (minutes) | Main Statistics | Time required to complete the processing (minutes) | Main Statistics |
| Logbook- VMSMatching | 5 | 60.3% of VMS tracks with matching in LB database | 30 | 57.2% of VMS tracks with matching in LB database | |
| Predict Métier: DataPreparation | 10 | / | 50 | / | |
| Predict Métier:Training ANN | 3 | Prediction completed for all the tracks without métiers by LB database | 3 | Prediction completed for all the trackswithout métiers by LB database | |
| Find Fishing Points | 30 | / | 45 | / | |
| Subtotal | ∼50 (∼1 hours) | ∼128 (∼2 hours) | |||
| Data Output | Tool | Time required to complete the processing (minutes) | Main Statistics | Time required to complete the processing (minutes) | Main Statistics |
| Gridding & Mapping(for each métier) | 2 | / | 9 | / | |
| DCF Indicators(for each métier) | <1 | / | <1 | / | |
| Trawled Area Viewer | Untested | / | / | / | |
| Subtotal | ∼3 | ∼10 | |||
| Complete data flow | Total | ∼600 minutes (∼10 hours) | ∼3180 minutes (∼53 hours/2.5 days) | ||
These data was measured on three sample datasets (two for VMS and one for Logbook).
$
The Fast & Heavy Algorithm was tested on a different personal computer with 16 Gb of Ram.