如何将子图中具有值和 x-y 位置的列绘制为颜色网格

问题描述 投票:0回答:1

在 R 中,我将执行以下操作来制作一个面网格,每个面都有一个栅格图:

# R Code

DF <- data.frame(expand.grid(seq(0, 7), seq(0, 7), seq(0, 5)))
names(DF) <- c("x", "y", "z")
DF$I <- runif(nrow(DF), 0, 1)
#      x y z          I
#   1: 0 0 0 0.70252977
#   2: 1 0 0 0.74346071
#  ---                 
# 383: 6 7 5 0.93409337
# 384: 7 7 5 0.14143277
library(ggplot2)
ggplot(DF, aes(x = x, y = y, fill = I)) + 
  facet_wrap(~z, ncol = 3) +
  geom_raster() + 
  scale_fill_viridis_c() +
  theme(legend.position = "bottom") # desired legend position should be bottom

我怎样才能在Python中做到这一点(使用matplotlib,可能还使用seaborn)?我使用以下代码进行了尝试,但在使用

plt.imshow
尝试绘制图像时遇到了问题。由于必须为
plt.imshow
重新调整数据,我想我需要为
g.map
自定义绘图函数。我尝试了几件事,但在轴或颜色以及在自定义绘图函数中使用数据时遇到了问题。

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import itertools

df = pd.DataFrame(list(itertools.product(range(8), range(8), range(6))), 
                  columns=['x', 'y', 'z'])
# order of values different than in R, but that shouldn't matter for plotting
df['I'] = np.random.rand(df.shape[0])
#      x  y  z         I
# 0    0  0  0  0.076338
# 1    0  0  1  0.148386
# 2    0  0  2  0.481053
# ..  .. .. ..       ...
# 382  7  7  4  0.144188
# 383  7  7  5  0.700624
g = sns.FacetGrid(df, col='z', col_wrap=2, height=4, aspect=1)
g.map(plt.imshow, color = 'I') # <- plt.imshow does not work here. 
# How can this be corrected (probably with a custom plot function)?
plt.show()
python matplotlib ggplot2 seaborn raster
1个回答
0
投票
import seaborn as sns
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# sample data
df = pd.DataFrame(list(itertools.product(range(8), range(8), range(6))), 
                  columns=['x', 'y', 'z'])
np.random.seed(20231116)  # for reproducible data
df['I'] = np.random.rand(df.shape[0])

# create the figure and axes
fig, axes = plt.subplots(2, 3, figsize=(15, 10), sharex=True, sharey=True)

# flatten the axes into a 1d array for easy access
axes = axes.flat

# add a separate axes for the colorbar
cbar_ax = fig.add_axes([0.3, .03, .4, .03])

# enumerate is specifically for adding the colorbar
# zip each group of 'z' data to the appropriate axes
for i, (ax, (z, data)) in enumerate(zip(axes, df.groupby('z'))):

    # pivot data into the correct shape for heatmap
    data = data.pivot(index='y', columns='x', values='I')

    # plot the heatmap
    sns.heatmap(data=data, cmap='viridis', ax=ax, cbar=i == 0, vmin=df.I.min(), vmax=df.I.max(),
                cbar_ax=None if i else cbar_ax, cbar_kws=dict(location="bottom"))

    # add a title
    ax.set(title=f'Z: {z}')

    # invert the yaxis to match the OP
    ax.invert_yaxis()

data
对于
z: 5

x         0         1         2         3         4         5         6         7
y                                                                                
0  0.488408  0.855913  0.339374  0.452842  0.510380  0.690491  0.448773  0.500916
1  0.273653  0.561840  0.860269  0.387470  0.170281  0.718488  0.256749  0.463527
2  0.546085  0.093934  0.273339  0.503968  0.063212  0.537974  0.867814  0.135719
3  0.071505  0.792265  0.919784  0.559663  0.733996  0.032003  0.475792  0.690789
4  0.474310  0.265576  0.841875  0.496676  0.603356  0.328808  0.039460  0.461778
5  0.439142  0.119253  0.842653  0.155213  0.798092  0.093709  0.899745  0.927067
6  0.548373  0.259983  0.295939  0.700694  0.040197  0.679880  0.153048  0.328768
7  0.216977  0.176777  0.238436  0.610802  0.705161  0.614877  0.813430  0.527120
© www.soinside.com 2019 - 2024. All rights reserved.