dpabon / pylandstats

Computing landscape metrics in the Python ecosystem

PyLandStats

Open-source library to compute landscape metrics in the Python ecosystem (NumPy, pandas, matplotlib...)

Citation: Bosch M. 2019. "PyLandStats: An open-source Pythonic library to compute landscape metrics". PLOS ONE, 14(12), 1-19. doi.org/10.1371/journal.pone.0225734

Features

• Read GeoTiff files of land use/cover:

  import pylandstats as pls
  
  ls = pls.Landscape("../data/processed/veveyse-AS18_4.tif")
  ls.plot_landscape(legend=True)

  

• Compute pandas data frames of landscape metrics at the patch, class and landscape level:

  class_metrics_df = ls.compute_class_metrics_df(
      metrics=["proportion_of_landscape", "edge_density", "euclidean_nearest_neighbor_mn"]
  )
  class_metrics_df

  class_val	proportion_of_landscape	edge_density	euclidean_nearest_neighbor_mn
  1	7.749572	19.102211	309.244705
  2	56.271868	50.599270	229.079970
  3	33.946252	38.167200	253.299859
  4	2.032308	3.722177	552.835154

• Analyze the spatio-temporal evolution of landscapes:

  import matplotlib.pyplot as plt
  
  input_filepaths = [
      "../data/processed/veveyse-AS97R_4.tif",
      "../data/processed/veveyse-AS09R_4.tif",
      "../data/processed/veveyse-AS18_4.tif",
  ]
  
  sta = pls.SpatioTemporalAnalysis(input_filepaths, dates=["1992", "2004", "2012"])
  sta.plot_metric("contagion")

  

• Zonal analysis of landscapes

See the documentation and the pylandstats-notebooks repository for a more complete overview.

Installation

The easiest way to install PyLandStats is with conda:

$ conda install -c conda-forge pylandstats

which will install PyLandStats and all of its dependencies. Alternatively, you can install PyLandStats using pip:

$ pip install pylandstats

Nevertheless, note that in order to define zones by vector geometries in ZonalAnalysis, or in order to use the the BufferAnalysis and SpatioTemporalBufferAnalysis classes, PyLandStats requires geopandas, which cannot be installed with pip. If you already have the dependencies for geopandas installed in your system, you might then install PyLandStats with the geo extras as in:

$ pip install pylandstats[geo]

and you will be able to use the aforementioned features (without having to use conda).

Development install

To install a development version of PyLandStats, you can first use conda to create an environment with all the dependencies and activate it as in:

$ conda create -n pylandstats -c conda-forge geopandas matplotlib-base rasterio scipy openblas
$ conda activate pylandstats

and then clone the repository and use pip to install it in development mode

$ git clone https://github.com/martibosch/pylandstats.git
$ cd pylandstats/
$ pip install -e .

Acknowledgments

• The computation of the adjacency matrix in transonic has been implemented by Pierre Augier (paugier)
• Several information theory-based metrics from Nowosad and Stepinski [1] were added by achennu
• With the support of the École Polytechnique Fédérale de Lausanne (EPFL)
• The Corine Land Cover datasets used for the test datasets were produced with funding by the European Union

References

1. Nowosad, J., & Stepinski, T. F. (2019). Information theory as a consistent framework for quantification and classification of landscape patterns. Landscape Ecology, 34(9), 2091-2101.