Plot a dictionary entrys as 3D Plot

Question

I have a dictionary, where each entry represents a pandas dataframe. The dictionary has 100 entrys.

The dataframes represent Absorption measurements over a given distance along the X-axis, keeping the Y-axis position fixed. Each dataframe corresponds to a different Y position. Those steps between the Y positions are fixed (100µm).

I want to plot those DataFrames next to each other, in a single plot, considering also the Y-axis. Therefore creating a 3D plot.

Up until now, I only can plot all the DataFrames in one single 2D plot.

The picture attached shows the idea.

Answer 1

See the code example below, which includes some test data.

The code assumes there's a common x-axis across all Absorption entries. If each absorption list is recorded over different x, then an additional interpolation step would need to be added to map all the data onto a common axis.

import pandas as pd
import numpy as np
from mpl_toolkits import mplot3d
from matplotlib import pyplot as plt

#Synthesise data
# data_dict contains 100 dataframes, indexed data_dict[0] to data_dict[99].
# Each dataframe has 2 columns: Absorption, and the x values at which the measurement was made
np.random.seed(0)

num_x_points = 200
x_axis = np.linspace(0, 2*np.pi / 2, num_x_points)
data_dict = {}
for i in range(100):
    abs_data = (np.random.uniform(0.9, 1.1)
                - np.cos(x_axis + np.random.uniform(0, i*np.pi/180))
                ) / (0.1*i + 10)
    
    data_dict[i] = pd.DataFrame({
        'Absorption': abs_data,
        'x_axis': x_axis
    })

#Determine number of y points
num_y = len(data_dict)
#Create y axis (in um)
y_step = 100e-6
y_axis = 1e6 * np.linspace(0, num_y * y_step, num_y)

#Compile absorption lines into a single matrix
abs_lines = np.empty((num_y, len(x_axis)))
for y_idx, df in enumerate(data_dict.values()):
    abs_lines[y_idx, :] = df.Absorption.to_numpy()

#Create a meshgrid for plotting in 3D
grid_x, grid_y = np.meshgrid(x_axis, y_axis)

#You can view the data either as lines with solid colours,
# or as points coloured by absorption
view_as_scatter = True

#Make 3d figure
figsize = 8
ax = plt.figure(figsize=[figsize] * 2).add_subplot(projection='3d')
#adjust the viewing angles
ax.view_init(elev=20, azim=240, roll=0)

if view_as_scatter:
    ax.scatter3D(grid_x, grid_y, abs_lines, c=abs_lines, marker='.', s=3, cmap='brg')
else:
    for y_idx, df in enumerate(data_dict.values()):
        ax.plot3D(grid_x[y_idx, :], grid_y[y_idx, :], df.Absorption)
    
ax.set_xlabel('x')
ax.set_ylabel('y (um)')
ax.set_zlabel('absorption')