sampley 0.0.7

sample survey data

sampley: sample survey data

Description

The sampley package serves to sample survey data (hence 'sampley': 'sample' + 'survey'). By 'survey data', we refer principally to systematic visual survey data, nevertheless, the sampley package may also be applicable to other kinds of survey data. By 'sample', we mean process those data to produce distinct samples that can then be input to a model.

Data can be processed by one or more of three approaches referred to here as the grid, segment, and point approach.
The grid approach consists of overlaying a grid, typically rectangular or hexagonal, onto the study area and allocating detections and, optionally, survey effort to the cells that they lie within. Additionally, data may be allocated to temporal periods. Each cell within a given period then serves as a sample.
The segment approach involves taking sections of continuous, uniform survey effort and cutting them into segments of standardised lengths which serve as samples.
The point approach consists of using the detections, or a subset thereof, as presences and then sampling absences from absence zones (i.e., areas that were surveyed but where the species was not detected).

For more information, please consult the User Manual (available on the GitHub repo at: https://github.com/JSBigelow/sampley/blob/main/sampley%20-%20User%20Manual.pdf).

Installation

pip install sampley

Import

For basic utilisation of sampley, run:
from sampley import *

To access the underlying functions, run:
from sampley.functions import *

User Manual

A user manual containing more detailed information is available on GitHub at: https://github.com/JSBigelow/sampley/blob/main/sampley%20-%20User%20Manual.pdf

Example usage

Several exemplars illustrating how to use sampley are available on GitHub at: https://github.com/JSBigelow/sampley/tree/main/exemplars

See the Introduction to sampley exemplars (intro.ipynb) for more information (https://github.com/JSBigelow/sampley/blob/main/exemplars/intro.ipynb)

License

MIT