Spaces:
Sleeping
Sleeping
Download docs/predict.html from sammmmmmm25/marine-tracker: direct link, hf CLI and curl.
- Browser
- Download file 6.96 kB
-
https://huggingface.co/spaces/sammmmmmm25/marine-tracker/resolve/main/docs/predict.html
- Command line
-
hf download hf://spaces/sammmmmmm25/marine-tracker/docs/predict.html
-
curl -L -o predict.html https://huggingface.co/spaces/sammmmmmm25/marine-tracker/resolve/main/docs/predict.html
6.96 kB
| <html lang="en"> | |
| <head> | |
| <meta charset="utf-8"> | |
| <title>Python: module predict</title> | |
| </head><body> | |
| <table class="heading"> | |
| <tr class="heading-text decor"> | |
| <td class="title"> <br><strong class="title">predict</strong></td> | |
| <td class="extra"><a href=".">index</a><br><a href="file:/Users/samuelveliveli/Downloads/marine-tracker/src/predict.py">/Users/samuelveliveli/Downloads/marine-tracker/src/predict.py</a></td></tr></table> | |
| <p><span class="code">Near-term movement projection: two honesty-first tiers.<br> | |
| <br> | |
| `heading` (default): estimate current bearing and speed from the last N<br> | |
| fixes, project forward linearly, and grow an uncertainty cone whose width is<br> | |
| grounded in that individual's own historical bearing/speed variance -- not an<br> | |
| arbitrary constant. This is the credible baseline.<br> | |
| <br> | |
| `markov` (optional upgrade): discretize the individual's own historical<br> | |
| bearing changes into direction bins, build a transition model from its own<br> | |
| track, and simulate a small ensemble of forward paths. The output is a<br> | |
| spread of plausible paths, not one line.<br> | |
| <br> | |
| Neither tier should be read as a forecast of certainty. Both are labeled<br> | |
| "projected range of motion" by the callers in analysis.py/app.py.</span></p> | |
| <p> | |
| <table class="section"> | |
| <tr class="decor pkg-content-decor heading-text"> | |
| <td class="section-title" colspan=3> <br><strong class="bigsection">Modules</strong></td></tr> | |
| <tr><td class="decor pkg-content-decor"><span class="code"> </span></td><td> </td> | |
| <td class="singlecolumn"><table><tr><td class="multicolumn"><a href="numpy.html">numpy</a><br> | |
| </td><td class="multicolumn"><a href="pandas.html">pandas</a><br> | |
| </td><td class="multicolumn"></td><td class="multicolumn"></td></tr></table></td></tr></table><p> | |
| <table class="section"> | |
| <tr class="decor functions-decor heading-text"> | |
| <td class="section-title" colspan=3> <br><strong class="bigsection">Functions</strong></td></tr> | |
| <tr><td class="decor functions-decor"><span class="code"> </span></td><td> </td> | |
| <td class="singlecolumn"><dl><dt><a name="-bearing_deg"><strong>bearing_deg</strong></a>(lat1, lon1, lat2, lon2)</dt><dd><span class="code">Initial great-circle bearing in degrees (0=north, 90=east) from point 1 to point 2.</span></dd></dl> | |
| <dl><dt><a name="-circular_mean_deg"><strong>circular_mean_deg</strong></a>(bearings)</dt><dd><span class="code">Mean of a list of compass bearings (degrees), handling the 0/360 wraparound.</span></dd></dl> | |
| <dl><dt><a name="-circular_std_deg"><strong>circular_std_deg</strong></a>(bearings)</dt><dd><span class="code">Circular standard deviation of compass bearings (degrees), handling wraparound.</span></dd></dl> | |
| <dl><dt><a name="-destination_point"><strong>destination_point</strong></a>(lat, lon, bearing, distance_km)</dt><dd><span class="code">Great-circle destination point given a start point, bearing (deg), and distance (km).</span></dd></dl> | |
| <dl><dt><a name="-predict"><strong>predict</strong></a>(individual_df, mode='heading', **kwargs)</dt><dd><span class="code">Dispatch to predict_heading (default) or predict_markov.</span></dd></dl> | |
| <dl><dt><a name="-predict_heading"><strong>predict_heading</strong></a>(individual_df, window=10, horizon_hours=24, step_hours=1)</dt><dd><span class="code">Linear heading/speed projection with a variance-grounded expanding uncertainty cone.<br> | |
| <br> | |
| Uses the last `window` segments to estimate current heading and speed,<br> | |
| and the individual's *entire* track history to estimate bearing/speed<br> | |
| variance (so the cone reflects how erratically this individual actually<br> | |
| moves, not an arbitrary assumption). Returns a dict with:<br> | |
| - centerline: list of (timestamp, lat, lon) for the projected path<br> | |
| - cone_left / cone_right: matching boundary lat/lon lists<br> | |
| - summary: short plain-language description<br> | |
| - heading_deg, speed_kmh, bearing_std_deg, speed_std_deg (for transparency)</span></dd></dl> | |
| <dl><dt><a name="-predict_markov"><strong>predict_markov</strong></a>( | |
| individual_df, | |
| n_bins=16, | |
| horizon_hours=24, | |
| step_hours=1, | |
| n_runs=50, | |
| random_state=None | |
| )</dt><dd><span class="code">Sample an ensemble of forward paths from a bearing-transition model built on this individual's own history.<br> | |
| <br> | |
| Returns a dict with `paths`: a list of n_runs paths, each a list of<br> | |
| (timestamp, lat, lon), plus a `summary`. Communicates uncertainty as a<br> | |
| spread of simulated paths rather than a single predicted line.</span></dd></dl> | |
| <dl><dt><a name="-track_segments"><strong>track_segments</strong></a>(individual_df)</dt><dd><span class="code">Return per-consecutive-fix bearing (deg) and speed (km/h) for one individual's sorted track.<br> | |
| <br> | |
| Expects individual_df sorted by timestamp with location_lat/location_long columns.<br> | |
| Returns a DataFrame of len(individual_df) - 1 rows with columns bearing, speed_kmh, dt_hours.</span></dd></dl> | |
| </td></tr></table><p> | |
| <table class="section"> | |
| <tr class="decor data-decor heading-text"> | |
| <td class="section-title" colspan=3> <br><strong class="bigsection">Data</strong></td></tr> | |
| <tr><td class="decor data-decor"><span class="code"> </span></td><td> </td> | |
| <td class="singlecolumn"><strong>EARTH_RADIUS_KM</strong> = 6371.0088</td></tr></table> | |
| </body></html> |