来自元组列表的 Python 热图答案

【问题标题】：Python heatmaps from list of tuples来自元组列表的 Python 热图
【发布时间】：2020-05-21 03:25:09
【问题描述】：

我有一个看起来像这样的数据：

# X,Y,Value
(380.57721129859806, 625.4295833013282, 1610.7896478197865)
(595.9181549398772, -309.2877476992412, 2153.4213188808317)
(85.88733459241405, 652.1788114374065, 497.6607201161437)
(247.44098377595287, -619.5256146069361, 1283.9468394229907)
(-259.092425954383, -383.41841661290914, 1850.040750164471)
(-431.58095056080657, -385.88458762039073, 1697.8866748485123)
(-469.9205503612537, -631.0749916983557, 2062.3719844791462)
(538.9858923744944, -207.61857693940544, 2309.0439437122927)
(291.8537762055346, -332.97650146280097, 1095.5209433044436)
(-90.17989357135775, 253.36425453647644, 1347.6315490796333)

X,Y 的范围是 -700 到 700，value 的范围是 1 到 3000。我想使用该数据来创建热图。这是我正在使用的代码：

from io import BytesIO
from PIL import Image
import matplotlib.pyplot as plt
import numpy as np

import scipy.ndimage.filters as filters
from matplotlib.colors import LinearSegmentedColormap

if __name__ == "__main__":
    w = 760
    h = 760

    data = np.zeros(h * w)
    data = data.reshape((h, w))

    for x in range(120, 123):
        for y in range(120, 123):
            data[x][y] = 2674

    for x in range(100, 103):
        for y in range(100, 103):
            data[x][y] = 1000

    data = filters.gaussian_filter(data, sigma=15)

    c_map = plt.cm.get_cmap("jet")
    transparent_jet = c_map(np.arange(c_map.N))
    transparent_jet[:, -1] = np.linspace(0, 1, c_map.N)
    cm = LinearSegmentedColormap.from_list('transparent_jet', transparent_jet)

    img_map = Image.new("RGBA", (760, 760), color="black")
    tmp = BytesIO()
    plt.imsave(tmp, data, cmap=cm)
    tmp.seek(0)
    Image.Image.alpha_composite(img_map, Image.open(tmp))
    img_map.save("heatmap.png")

如何改用浮点 X、Y 和值？

【问题讨论】：

你能用plt.scatter 用各自的颜色分散点吗？

标签： python numpy matplotlib scipy

【解决方案1】：

我只需要使用histogram2d。

from io import BytesIO
from PIL import Image
import matplotlib.pyplot as plt
import numpy as np

import scipy.ndimage.filters as filters
from matplotlib.colors import LinearSegmentedColormap

if __name__ == "__main__":
    c_map = plt.cm.get_cmap("jet")
    transparent_jet = c_map(np.arange(c_map.N))
    transparent_jet[:, -1] = np.linspace(0, 1, c_map.N)
    cm = LinearSegmentedColormap.from_list('transparent_jet', transparent_jet)

    h_x = np.random.uniform(-250, 250, 10000)
    h_y = np.random.uniform(-250, 250, 10000)
    h_w = np.random.uniform(1, 3000, 10000)

    h, _, _ = np.histogram2d(h_x, h_y, weights=h_w, bins=(500, 500))
    h = filters.gaussian_filter(h, sigma=6)

    img_map = Image.new("RGBA", (500, 500), color="black")
    tmp = BytesIO()
    plt.imsave(tmp, h, cmap=cm)
    tmp.seek(0)
    Image.Image.alpha_composite(img_map, Image.open(tmp))
    img_map.show()

结果图片：

【讨论】：