Welcome to igclib’s documentation¶

Introduction¶

igclib is a Python module and a command line tool designed for the analysis of paragliding competitions. Its original purpose is to serve as backend to Task Creator.

Installation¶

This package will be hosted on PyPI when it reaches its beta version. For the moment, you can install it by downloading the source from GitHub.

git clone https://github.com/teobouvard/igclib.git
cd igclib
make install

Examples¶

Python Examples¶

Build a race and save it to disk¶

from igclib.core.race import Race

r =  Race(tracks='tracks/', task='task.xctsk')
r.save('race.pkl')

Load a race and save it as json¶

from igclib.core.race import Race

r =  Race(path='race.pkl')
r.save('race.json')

Get a task optimization info and save it as json¶

from igclib.core.task import Task

t = Task('task.xctsk')
t.save('optimized.json')

Get all available tasks from a provider on standard output¶

from igclib.crawlers.task_crawler import TaskCrawler

crawler = TaskCrawler(provider='PWCA', year=2015)
crawler.save('-')

Command line Examples¶

Build a race and save it to disk¶

igclib race --task task.xctsk --flights tracks_dir/ --output race.pkl

Get a task optimization info in json¶

igclib optimize --task task.xctsk --progress ratio

Get all available tasks from a provider¶

igclib crawl --provider PWCA --year 2015

Architecture¶

Soon

Python API¶

Race¶

class igclib.core.race.Race(tracks=None, task=None, validate=True, path=None, progress='gui')[source]¶

You can create a Race instance in two different ways :

Passing a tracks and a task, which creates a new Race object and validates all pilot flights.
```
>>> r =  Race(tracks='tracks/', task='task.xctsk')
```
Passing a path to a previously saved Race, loading the saved instance (much faster than re-validating flights).
```
>>> r =  Race(path='race.pkl')
```

Keyword Arguments:
	tracks (str) – A path to a directory or a zip file containing IGC tracks. task (str) – A path to the task file or a base64 representation of the task. path (str) – The path of a previously saved Race instance.

n_pilots¶

The number of pilots in the Race.

Type:	int

flights¶

A collection of Flights indexed by pilot ID.

Type:	dict [str, Flight]

task¶

The Task instance of the Race.

Type:	Task

get_pilot_features(pilot_id, start=None, stop=None)[source]¶

Extracts pilot features

Keyword Arguments:
Parameters:	pilot_id (str) – The pilot identifier used as key in self.flights
	start (time, optional) – Lower bound of the retrieved features (default) stop (time, optional) – Upper bound of the retrieved features
Raises:	`KeyError` – if pilot_id is not a key of self.flights dictionnary
Returns:	The pilot features from start to stop
Return type:	PilotFeatures

parse_flights(tracks)[source]¶

Populates flights attribute by parsing each igc file in tracks.

Parameters:	tracks (str) – Path to a directory or a zip file containing the igc files

pilot_schema(pilot_id, output)[source]¶

In dev !

Parameters:	pilot_id (str) – ID of the pilot being watched

serialize()[source]¶: Serializes the race object to be written to a JSON file

validate_flights()[source]¶: Computes the validation of each flight on the race

Flight¶

class igclib.core.flight.Flight(igc_file)[source]¶: Class representing a Flight

Task¶

class igclib.core.task.Task(task)[source]¶

Creates a Task object. The way this is done is really bad, and any help would be appreciated. As of now, this class checks if the input task (file or b64) “fits” in each parser implemented. When a parser does not raise a KeyError, the task is assumed to be of this format and its attributes are copied back to the Task object. I feel that there is a smarter way to do this, probably with inheritance. However, task format is not known in advance so trying each format seems like the easiest thing to do.

Parameters:	task (str) – Path to or base64 representation of a task file.
Raises:	`NotImplementedError` – If the task could not be parsed.

Pilot Features¶

class igclib.core.pilot_features.GroupRelation(pilot_id, snapshot)[source]¶

Convention : If a GroupRelation feature is > 0, it means that the original pilot is in a better position than the other pilot in the group. This means that :

delta_altitude > 0 : original pilot is higher
delta_distance > 0 : original pilot closer to goal

other_pilot_id¶

A list of IDs pilots being compared

Type:	list [str]

delta_altitude¶

The altitude differences.

Type:	list [float]

delta_distance¶

The goal distances differences.

Type:	list [float]

class igclib.core.pilot_features.PilotFeatures(pilot_id, timestamp, snapshot)[source]¶

pilot_id¶

The pilot’s ID

Type:	str

timestamp¶

The timestamp associated with the features.

Type:	time

position¶

The current position of the pilot.

Type:	GeoPoint

group_relation¶

The current position of the pilot.

Type:	GroupRelation

Geography¶

This module contains the base classes representing a GeoPoint, a Turnpoint and an Opti. It also provides convenience wrappers around C extension function calls. These wrappers allow for more concise function calls to improve readability.

class igclib.geography.geo.Arc(lat, lon, p1=None, p2=None, radius=None)[source]¶

class igclib.geography.geo.GeoPoint(lat=None, lon=None, altitude=None, record=None, status=None)[source]¶

class igclib.geography.geo.Opti(dist, legs, points, angles)[source]¶

class igclib.geography.geo.Point(*args)[source]¶

class igclib.geography.geo.Turnpoint(lat, lon, radius=None, altitude=None, name=None, desc=None, role=None, first_tag=None)[source]¶

Crawlers¶

Command line manual¶

Usage¶

igclib COMMAND [OPTIONS] ARGS

Commands¶

optimize

    Get the optimized version of a task.
    Required arguments :
        --task
        --output

replay

    Creates a race without validating flights.
    Required arguments :
        --task
        --flights
        --output

race

    Creates a race with validated flights.
    Required arguments :
        --task and --flights or --path
        --output

crawl

    Crawls a provider for all events hosted in the specified year.
    Optional arguments :
        --provider [ PWCA | FFVL ] (defaults to PWCA)
        --year (defaults to current year)

watch

    Get the features of a pilot on a flight.
    Required arguments :
        --path
        --pilot
        --output

Options¶

--progress [ gui | ratio | silent ] (defaults to gui)

    Defines the way progress is displayed to the user.
    * gui displays a progress bar
    * ratio displays a percentage on each line
    * silent runs quietly

Files specifications¶

Soon