pygieons documentation

What is pygieons?

The name (you would anyway ask 😉): Python GIS and Earch Observation Networks

As described elsewhere, this website is largely generated automatically by using a Python tool called pygieons. This tool auto-generates content regarding the listed packages and adds them to the website either as tables or interactive Network visualizations. pygieons is currently focused on parsing information about Python GIS and Earth Observation libraries, but it is a rather generic tool, with modest modifications it would be possible to parse and visualize information from any collection of libraries available on PyPi.

pygieons is largely based on pandas which is used for data manipulation, while pyvis library is used to visualize the connections between libraries as an interactive network of links. In addition, under the hood libraries such as requests, beautifulsoup4 and pypistats are used to extract relevant information about the listed libraries from the PyPi web pages.

Below, we demonstrate some basic functionalities that can be used to auto-generate content using pygieons.

Basic usage

Initializing the tool

To generate content with pygieons, you need to first initialize the tool with Ecosystem() object which accepts a parameter plot_type that can be used to define which category of libraries we want to plot, and parameter log which can be used to disable logging to the screen. Below is the slightly modified output from help(Ecosystem) command showing information about how the tool can be initialized:

class Ecosystem(builtins.object)
 |  Ecosystem(plot_type='all', keep_all=False, log=True)
 |        
 |      Parameters
 |      ==========
 |      
 |      plot_type : str
 |          This can be used to specify if you want to plot all packages in same figure, or only vector or raster.
 |          Possible values are: "all", "vector", "raster", "generic", "vector+generic", "raster+generic".
 |      keep_all : bool
 |          If True, also packages that are not available from PyPi or which are not maintained will be kept.
 |      log: bool
 |          If True, will print messages during process.

• Let’s initialize the tool and configure it for plotting raster libraries.

from pygieons import Ecosystem

# Initialize the tool 
e = Ecosystem(plot_type="raster", log=True)

Generating tables

To generate a table showing various information about the libraries, you can use method prepare_table(), which autogenerates the table contents based on the packages listed in ecosystem_connections.py and ecosystem_pkgs.py files. The tool will print out information about the progress if log was specified as True:

# Generate the content
tbl = e.prepare_table()

Find out the number of monthly downloads from PyPi for the libraries ..

Extract project URLs ..

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[2], line 2
      1 # Generate the content
----> 2 tbl = e.prepare_table()

File ~/checkouts/readthedocs.org/user_builds/pythongis-ecosystem/envs/stable/lib/python3.10/site-packages/pygieons/pygieons.py:496, in Ecosystem.prepare_table(self, cols)
    493 self.nodes = prepare_html_links_and_badges(self.nodes)
    495 # Prepare table
--> 496 return Table(prepare_table_plot(self.nodes, cols))

File ~/checkouts/readthedocs.org/user_builds/pythongis-ecosystem/envs/stable/lib/python3.10/site-packages/pygieons/pygieons.py:300, in prepare_table_plot(nodes, cols)
    296 if "Documentation" in cols:
    297     align_center_cols.append("Documentation")
    299 return HTML(nodes[cols].style.set_properties(subset=align_center_cols, **{"text-align": "center"})
--> 300             .hide_index()
    301             .to_html()
    302             )

AttributeError: 'Styler' object has no attribute 'hide_index'

• You can visualize the table by calling .show() of the returned Table object:

tbl.show()

Analysis / modelling (count=9)

Name	Homepage	Info	License	PyPi version	PyPi downloads (monthly)	Conda-forge version	Conda-forge downloads	Conda-forge latest release
gstools	🏠	ⓘ			2,063
pykrige	🏠	ⓘ			18,550
pysheds	🏠	ⓘ			921
pyspatialml	🏠	ⓘ	NA		95	NA	NA	NA
rasterstats	🏠	ⓘ			61,296
richdem	🏠	ⓘ			2,129
scikit-learn	🏠	ⓘ			33,204,319
spyndex	🏠	ⓘ			356
xarray-spatial	🏠	ⓘ			2,074

Core / data structures (count=8)

Name	Homepage	Info	License	PyPi version	PyPi downloads (monthly)	Conda-forge version	Conda-forge downloads	Conda-forge latest release
affine	🏠	ⓘ			673,504
iris	🏠	ⓘ			951
rasterio	🏠	ⓘ			755,843
rio-cogeo	🏠	ⓘ			12,869
rioxarray	🏠	ⓘ			36,151
sarpy	🏠	ⓘ			943
sarsen	🏠	ⓘ			54
xarray	🏠	ⓘ			1,705,385

Data extraction / processing (count=21)

Name	Homepage	Info	License	PyPi version	PyPi downloads (monthly)	Conda-forge version	Conda-forge downloads	Conda-forge latest release
earthengine-api	🏠	ⓘ			36,212
easystac	🏠	ⓘ			110
eemont	🏠	ⓘ			700
lidar	🏠	ⓘ			283
odc-stac	🏠	ⓘ			986
planetary-computer	🏠	ⓘ			2,795
pymap3d	🏠	ⓘ			41,451
pyrosar	🏠	ⓘ			491
pystac	🏠	ⓘ			32,119
pystac-client	🏠	ⓘ			9,842
radiant-mlhub	🏠	ⓘ			5,836
rio-hist	🏠	ⓘ	NA		364	NA	NA	NA
rio-mucho	🏠	ⓘ			8,847
rio-tiler	🏠	ⓘ			8,237
salem	🏠	ⓘ			5,834
satpy	🏠	ⓘ			3,498
sentinelsat	🏠	ⓘ			12,394
stackstac	🏠	ⓘ			1,235
verde	🏠	ⓘ			731
xarray-sentinel	🏠	ⓘ			79
xyzservices	🏠	ⓘ			165,745

Visualization (count=3)

Name	Homepage	Info	License	PyPi version	PyPi downloads (monthly)	Conda-forge version	Conda-forge downloads	Conda-forge latest release
contextily	🏠	ⓘ			71,077
rio-color	🏠	ⓘ			8,170
xarray_leaflet	🏠	ⓘ			1,026

Generating Network visualization

To generate an interactive network visualization showing the connections between libraries, you can use method prepare_net(), which autogenerates the Network based on the packages listed in ecosystem_connections.py file. The tool will print out information about the progress if log was specified as True:

# Generate the content
net = e.prepare_net()

• You can visualize the network by calling .show() of the returned Network object

net.show()

• In addition, you can also save the Network visualization as a HTML file to your own computer by calling .save() method:

net.save("test.html")