MatPlotLib 繪圖沒有明確的邊界，奇怪的數據點

Question

我正在嘗試映射數據，但地圖輸出沒有明確的邊界，並且顯示的數據不應該是連續的。 下面的第一張地圖使用的數據與第二張地圖的確切代碼相似，所以我不知道出了什么問題。 我想知道是否有辦法格式化代碼，使情節在風格上與第一個相似。

import matplotlib.pyplot as plt
f, ax = plt.subplots(1, figsize=(12,6))
ax = states_00_14.plot(column='num_fires', cmap='OrRd',
legend=True, ax=ax)
lims = plt.axis('equal')
f.suptitle('US Wildfire count per state in 2000-2014')
ax.set_axis_off()

我對 python 和 matplotlib 很陌生，所以我基本上不知道我做錯了什么。 如果相關，我正在使用 Jupyter Notebook。 提前致謝！

Answer 1

第一個是等值線圖，它可以通過空間連接（合並）狀態幾何數據與野火位置數據（例如https://geopandas.org/docs/reference/api/geopandas.sjoin.html ）並使用 geopandas 繪制來生成（ https://geopandas.org/docs/user_guide/mapping.html ）。

第二張地圖是帶有圓形標記的散點圖，因此不清晰。

Answer 2

• 您沒有定義數據源。 已使用： https : //www.naturalearthdata.com/downloads/110m-culture-vectors/用於州界。 已使用https://data-nifc.opendata.arcgis.com/datasets/wfigs-wildland-fire-locations-full-history/explore?showTable=true作為野火源
• 這意味着現在必須對數據框進行地理處理，這很容易完成空間連接。 然后，這允許將狀態與火點相關聯。

# spatial join of fire locations to states
state_fire = gdf_fire.loc[fire_mask, fire_cols].sjoin(
    gdf2.loc[boundary_mask, boundary_cols]
)

• 關聯州后，可以匯總數據以獲取每個州每年的火災數量
• 首先使用plotly將這些可視化，因為這允許我為每年的幀設置動畫，以及使用懸停信息進行更簡單的調試
• 然后用matplotlib進行可視化。 重新創建您使用的相同格式的GeoDataFrame ，將年份聚合到 2000-2014，然后加入狀態多邊形進行繪圖
• 添加了edgecolor="black"以便清楚地標記邊緣。

import geopandas as gpd
import pandas as pd
import plotly.express as px
import requests
from pathlib import Path
from zipfile import ZipFile
import urllib
import requests

# get wild fire data..
# https://data-nifc.opendata.arcgis.com/datasets/wfigs-wildland-fire-locations-full-history/explore?showTable=true
gdf_fire = gpd.GeoDataFrame.from_features(
    requests.get(
        "https://opendata.arcgis.com/datasets/d8fdb82b3931403db3ef47744c825fbf_0.geojson"
    ).json()
)

# fmt: off
# download boundaries
url = "https://www.naturalearthdata.com/http//www.naturalearthdata.com/download/110m/cultural/ne_110m_admin_1_states_provinces.zip"
f = Path.cwd().joinpath(urllib.parse.urlparse(url).path.split("/")[-1])
# fmt: on

if not f.exists():
    r = requests.get(url, stream=True, headers={"User-Agent": "XY"})
    with open(f, "wb") as fd:
        for chunk in r.iter_content(chunk_size=128):
            fd.write(chunk)
    zfile = ZipFile(f)
    zfile.extractall(f.stem)

# load downloaded boundaries
gdf2 = gpd.read_file(str(f.parent.joinpath(f.stem).joinpath(f"{f.stem}.shp")))

# a bit of cleanup, data types and CRS
gdf_fire["FireDiscoveryDateTime"] = pd.to_datetime(gdf_fire["FireDiscoveryDateTime"])
gdf_fire = gdf_fire.set_crs("EPSG:4326")

# filters, US states and fires with a date...
boundary_cols = ["adm1_code", "iso_3166_2", "iso_a2", "name", "geometry"]
boundary_mask = gdf2["iso_a2"].eq("US")
fire_cols = ["OBJECTID", "FireDiscoveryDateTime", "geometry"]
# fire_mask = gdf_fire["FireDiscoveryDateTime"].dt.year.between(2010,2012)
fire_mask = ~gdf_fire["FireDiscoveryDateTime"].isna()

# spatial join of fire locations to states
state_fire = gdf_fire.loc[fire_mask, fire_cols].sjoin(
    gdf2.loc[boundary_mask, boundary_cols]
)

# summarize data by year and state
df_fires_by_year = (
    state_fire.groupby(
        boundary_cols[0:-1]
        + ["index_right", state_fire["FireDiscoveryDateTime"].dt.year],
        as_index=False,
    )
    .size()
    .sort_values(["FireDiscoveryDateTime", "index_right"])
)

# and finally visualize...
px.choropleth_mapbox(
    df_fires_by_year,
    geojson=gdf2.loc[boundary_mask, "geometry"].__geo_interface__,
    locations="index_right",
    color="size",
    animation_frame="FireDiscoveryDateTime",
    hover_name="name",
    hover_data={"index_right": False},
    color_continuous_scale="OrRd",
).update_layout(
    mapbox={
        "style": "carto-positron",
        "zoom": 2,
        "center": {"lat": 39.50, "lon": -98.35},
    },
    margin={"l": 0, "r": 0, "t": 0, "b": 0},
)

matplotlib

import matplotlib.pyplot as plt

# recreate dataframe in question.  Exclude Alaska and Hawaii as they mess up boundaries...
# further aggregate the defined years....
states_00_14 = gpd.GeoDataFrame(
    df_fires_by_year.loc[df_fires_by_year["FireDiscoveryDateTime"].between(2000, 2014)]
    .groupby("index_right", as_index=False)
    .agg({"size": "sum"})
    .merge(
        gdf2.loc[boundary_mask & ~gdf2["iso_3166_2"].isin(["US-AK","US-HI"])],
        left_on="index_right",
        right_index=True,
        how="inner",
    )
)

f, ax = plt.subplots(1, figsize=(12, 6))
ax = states_00_14.plot(column="size", cmap="OrRd", legend=True, ax=ax, edgecolor="black")
lims = plt.axis("equal")
f.suptitle("US Wildfire count per state in 2000-2014")
ax.set_axis_off()