open_cp.data¶

Encapsulates input data.

class open_cp.data.Point(x=0, y=0)¶

Bases: object

A simple 2 dimensional point class.

x¶

The x coordinate.

y¶

The y coordinate.

class open_cp.data.RectangularRegion(xmin=0, xmax=1, ymin=0, ymax=1)¶

Bases: object

Stores a rectangular region.

aspect_ratio¶

Height divided by width

grid_size(cell_width, cell_height=None)¶

Return the size of grid defined by this region.

Parameters:	cell_width – The width of each cell in the grid. cell_height – Optional. The height of each cell in the grid; defaults to a square grid where the height is the same as the width.
Returns:	(xsize, ysize) of the grid.

height¶

The height of the region: ymax - ymin

max¶

The pair (xmax, ymax)

min¶

The pair (xmin, ymin)

width¶

The width of the region: xmax - xmin

xmax¶

xmin¶

ymax¶

ymin¶

class open_cp.data.TimedPoints(timestamps, coords)¶

Bases: object

Stores a list of timestamped x-y coordinates of events.

Parameters:	timestamps – An array of timestamps (must be convertible to numpy.datetime64). coords – An array of shape (2,n) where n must match the number of timestamps.

bounding_box¶

The smallest (space) box containing all the data points.

Returns:	A RectangularRegion instance.

empty¶

True or False, do we have any events

events_before(cutoff_time=None)¶

Returns just the events with timestamps before (or equal to) the cutoff.

Parameters:	cutoff_time – End of the time period we’re interested in. Default is None which means return all the data.

static from_coords(timestamps, xcoords, ycoords)¶

Static constructor allowing you to pass separate arrays of x and y coordinates.

number_data_points¶

The number of events

time_deltas(time_unit=numpy.timedelta64(1, 'm'))¶

Returns a numpy array of floats, converted from the timestamps, starting from 0, and with the optional unit.

Parameters:	time_unit – The unit to measure time by. Defaults to 1 minute, so timestamps an hour apart will be converted to floats 60.0 apart. No rounding occurs, so there is no loss in accuracy by passing a different time unit.

time_range¶

Find the time range.

Returns:	A pair (start, end) of timestamps.

times_datetime()¶

Return an array of timestamps using the datetime.datetime standard library class. Useful for plotting with matplotlib, for example.

to_time_space_coords(time_unit=numpy.timedelta64(1, 'm'))¶

Returns a single numpy array [t,x,y] where the time stamps are converted to floats, starting from 0, and with the optional unit.

Parameters:	time_unit – The unit to measure time by. Defaults to 1 minute, so timestamps an hour apart will be converted to floats 60.0 apart. No rounding occurs, so there is no loss in accuracy by passing a different time unit.

xcoords¶

A one dimensional array representing the x coordinates of events.

ycoords¶

A one dimensional array representing the y coordinates of events.

open_cp.data.points_from_lon_lat(points, proj=None, epsg=None)¶

Converts longitude / latitude data into x,y coordinates using a projection. The module pyproj must be loaded, otherwise this does nothing.

Parameters:	points – A :class TimedPoints: instance of lon/lat data. proj – Optionally, a pyproj.Proj instance describing the projection. epsg – If no proj is given, this must be supplied. A valid EPSG projection reference. For example, 7405 is suitable for UK data. See http://spatialreference.org/ref/epsg/
Returns:	A TimedPoints instance of projected data with the same timestamps.