Note:

This library was originally created at Jakarto because we needed a single library to easily create las files from pandas arrays. We found using the old laspy api a bit unintuitive, so we created this wrapper. Since laspy 2.0.0, this has been greatly improved. Someone looking for the simplest way to read and write las files could find jaklas useful, but I would encourage reading laspy's documentation also.

jaklas

jaklas is a thin wrapper around laspy to make reading and writing las files as simple as possible.

The main use case is to write a pandas array to a las file in a single function call. The las file attributes (point offset, point scaling, file version, point format, etc.) are inferred depending on column names, datatype and point values.

The las writer supports any object implementing __getitem__ that has the correct field names.

Installation

pip install jaklas

Testing

git clone git@github.com:jakarto3d/jaklas.git
cd jaklas
pip install -r requirements-dev.txt
python -m pip install .
pytest

Usage

jaklas.write writes a pandas dataframe (or a dict) to a las file.

The dataframe must have either (case insensitive):

'x', 'y' and 'z' columns
or an 'xyz' column

and it can have other las attributes (case sensitive names taken from laspy):

gps_time
intensity
classification
red
green
blue
edge_of_flight_line
key_point
nir
number_of_returns
overlap
point_source_id
raw_classification
return_number
return_point_wave_location
scan_angle
scan_angle_rank
scan_direction_flag
scanner_channel
synthetic
user_data
wavepacket_index
wavepacket_offset
wavepacket_size
withheld
x_t
y_t
z_t

other column names will be written as extra dimensions.

Example

import jaklas
import pandas

data = {
    'gps_time': [0, 1.232, 2.543, 3.741],
    'intensity': [14578, 54236, 1425, 12543],
    'X': [456, 234, 567, 432],
    'Y': [10234, 10256, 10789, 10275],
    'Z': [10, 11, 12, 13],
}
dataframe = pandas.DataFrame(data)
filename = 'example.las'
jaklas.write(dataframe, filename)

Note the upper case 'X', 'Y' and 'Z' point data are the real coordinates, not the scaled int32 ones like in the las file.