Vizualized model¶

ABSESpy provides a basic visualization of the model. The visualization is based on the matplotlib, xarray, geopandas, and networkx library. The visualization is not perfect, but it can help to understand the model structure.

Grid world¶

In [ ]:

from abses import MainModel

model = MainModel()

from abses import MainModel model = MainModel()

[14:33:13][WARNING][nature] the nature's CRS has been changed to epsg:4326.

In [ ]:

# Create a spatial module for actors
module = model.nature.create_module(shape=(50, 40), name="world")

# Create a spatial module for actors module = model.nature.create_module(shape=(50, 40), name="world")

In [ ]:

from abses import Actor
import matplotlib.pyplot as plt
import seaborn as sns


class Triangle(Actor):
    color = "#FF7F3E"
    marker = "^"

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.test = self.random.normalvariate(60, 20)


class MyUser(Actor):
    marker = "o"
    color = "green"

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.test = self.random.normalvariate(78, 5)


new_agents = module.random.new(Triangle, size=200, replace=True)
new_users = module.random.new(MyUser, size=300, replace=True)

# Plot histogram of the 'test' attribute
values = model.actors.array("test")
fig, ax = plt.subplots(figsize=(6,4))
sns.histplot(values, bins=30, kde=False, ax=ax)
ax.set_title("Distribution of test attribute")
plt.show()

from abses import Actor import matplotlib.pyplot as plt import seaborn as sns class Triangle(Actor): color = "#FF7F3E" marker = "^" def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) self.test = self.random.normalvariate(60, 20) class MyUser(Actor): marker = "o" color = "green" def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) self.test = self.random.normalvariate(78, 5) new_agents = module.random.new(Triangle, size=200, replace=True) new_users = module.random.new(MyUser, size=300, replace=True) # Plot histogram of the 'test' attribute values = model.actors.array("test") fig, ax = plt.subplots(figsize=(6,4)) sns.histplot(values, bins=30, kde=False, ax=ax) ax.set_title("Distribution of test attribute") plt.show()

Out[ ]:

<Axes: xlabel='test', ylabel='Count'>

In [ ]:

# Scatter plot of actor positions by reading coordinates
import matplotlib.pyplot as plt
xs = [a.at.pos[0] for a in model.actors if hasattr(a.at, "pos")]
ys = [a.at.pos[1] for a in model.actors if hasattr(a.at, "pos")]
colors = [getattr(a.__class__, "color", "gray") for a in model.actors]
plt.figure(figsize=(6, 5))
plt.scatter(xs, ys, c=colors, alpha=0.9, s=20)
plt.title("Actor positions")
plt.xlabel("x"); plt.ylabel("y")
plt.show()

# Scatter plot of actor positions by reading coordinates import matplotlib.pyplot as plt xs = [a.at.pos[0] for a in model.actors if hasattr(a.at, "pos")] ys = [a.at.pos[1] for a in model.actors if hasattr(a.at, "pos")] colors = [getattr(a.__class__, "color", "gray") for a in model.actors] plt.figure(figsize=(6, 5)) plt.scatter(xs, ys, c=colors, alpha=0.9, s=20) plt.title("Actor positions") plt.xlabel("x"); plt.ylabel("y") plt.show()

Out[ ]:

<Axes: xlabel='x', ylabel='y'>

Real-world map¶

In this section, we will visualize the real-world map of the model.

In [ ]:

from abses import MainModel
from abses import load_data

sc = MainModel()
cropland = sc.nature.create_module(
    how="from_file",
    raster_file=load_data("farmland.tif"),
    apply_raster=True,
    name="cropland",
    attr_name="test",
)

# adding some Triangle actors and then plotting
actors = cropland.random.new(Triangle, size=100)
from abses.viz.plotting import plot_raster
plot_raster(cropland, attr="test")

from abses import MainModel from abses import load_data sc = MainModel() cropland = sc.nature.create_module( how="from_file", raster_file=load_data("farmland.tif"), apply_raster=True, name="cropland", attr_name="test", ) # adding some Triangle actors and then plotting actors = cropland.random.new(Triangle, size=100) from abses.viz.plotting import plot_raster plot_raster(cropland, attr="test")

[14:33:14][WARNING][nature] the nature's CRS has been changed to GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AXIS["Latitude",NORTH],AXIS["Longitude",EAST],AUTHORITY["EPSG","4326"]].
/Users/songshgeo/Documents/VSCode/ABSESpy/abses/nature.py:170: UserWarning: Converting PatchModule from crs EPSG:4326 to the crs of BaseNature - EPSG:4326. Please check your crs settings if this is unintended, or set `GeoSpace.warn_crs_conversion` to `False` to suppress this warning message.
  self.add_layer(module)

Out[ ]:

<Axes: title={'center': 'spatial_ref = 0'}, xlabel='x', ylabel='y'>

Shapefile Dataset¶

ABSESpy allows user to create Geo- Agents from GeoDataFrame object.

In [5]:

from abses import load_data
import geopandas as gpd

data = gpd.read_file(load_data("YR_cities.zip"))
data.head()

from abses import load_data import geopandas as gpd data = gpd.read_file(load_data("YR_cities.zip")) data.head()

Out[5]:

	City_ID	Ratio	area	Shaoefile_	Perfecture	Province_n	geometry
0	100	0.546595	3958.281245	128.0	C100	Henan	POLYGON ((113.38580 35.47040, 113.39200 35.464...
1	102	0.385113	15258.027194	170.0	C102	Henan	POLYGON ((112.03220 35.04700, 112.03260 35.045...
2	106	0.567869	4273.313471	197.0	C106	Henan	POLYGON ((115.19540 36.20580, 115.19700 36.203...
3	107	0.389969	9951.746614	215.0	C107	Henan	POLYGON ((111.96140 35.07840, 111.96400 35.076...
4	109	0.601268	8279.739196	290.0	C109	Henan	POLYGON ((113.67120 35.83740, 113.67660 35.835...

In [6]:

model2 = MainModel()
cities = model2.agents.new_from_gdf(
    data, unique_id="City_ID", attrs={"area": "area", "Province_n": "province"}
)
cities

model2 = MainModel() cities = model2.agents.new_from_gdf( data, unique_id="City_ID", attrs={"area": "area", "Province_n": "province"} ) cities

Out[6]:

<ActorsList: (59)Actor>

In [ ]:

# Plot cities using networkx from links or simply scatter centroids
import matplotlib.pyplot as plt

try:
    # If cities carry geometry, scatter centroids
    xs = [getattr(a, "geom").centroid.x for a in cities if hasattr(a, "geom")]
    ys = [getattr(a, "geom").centroid.y for a in cities if hasattr(a, "geom")]
    if xs and ys:
        plt.figure(figsize=(5,4))
        plt.scatter(xs, ys, s=20)
        plt.title("Cities (centroids)")
        plt.show()
    else:
        raise AttributeError
except Exception:
    # fallback to listing
    print(f"Loaded {len(cities)} city actors.")

# Plot cities using networkx from links or simply scatter centroids import matplotlib.pyplot as plt try: # If cities carry geometry, scatter centroids xs = [getattr(a, "geom").centroid.x for a in cities if hasattr(a, "geom")] ys = [getattr(a, "geom").centroid.y for a in cities if hasattr(a, "geom")] if xs and ys: plt.figure(figsize=(5,4)) plt.scatter(xs, ys, s=20) plt.title("Cities (centroids)") plt.show() else: raise AttributeError except Exception: # fallback to listing print(f"Loaded {len(cities)} city actors.")

In [ ]:

# Already visualized above; show count as simple output
len(cities)

# Already visualized above; show count as simple output len(cities)

Out[ ]:

<Axes: >