如何在 python pandas 中绘制/分散地理坐标

Question

I'm having a hard time plotting something in pandas the way I want it.

I have movement traces in a csv file that look like that:

NodeID | Time   | Lon              | Lat             # Line not in the CSV
0      | 38.665 |29564.86465677584 |37278.27065254189
0      | 64.29  |29529.86419382295 |37300.74058444612
0      | 80.74  |29511.18207467592 |37317.11012177728
1      | 166.3  |29593.54098394629 |37403.83872184437
1      | 188.98 |29622.25878085964 |37441.86538931914
1      | 219.33 |29658.04716892622 |37491.65280349273
.
.
.
20     | 566.3  |29593.54098394629 |37403.83872184437
20     | 888.98 |29622.25878085964 |37441.86538931914
30     | 919.33 |29658.04716892622 |37491.65280349273

I read the csv with:

df = pd.read_csv(path, delimiter=' ', 
                 names=["node", "time", "x","y"], header=None)

Now I want to plot every node with a different color. Ideally, every point should be connected to the next, but this is not necessary due to the close distance of the locations.

I tried it with:

df.plot(x='x', y='y', kind='scatter')
# or
df[NodeID:].plot(x='x', y='y', kind='scatter')
# or with matplotlib
ax = fig.subplots()
ax.plot(pddata['x'], pddata['y'], label='node ' + str(pddata['node'])) 

But it always looks like this:

Can someone give me a hint how to plot it the way I want it?

Answer 1

This can be done with the seaborn package, it can be quite helpful in plotting and works with pandas . Looking through its gallery as well as the python graph gallery the might also give some inspiration for your plots.

https://seaborn.pydata.org/ https://python-graph-gallery.com/

For your specific problem, with the dataframe set up, this hopefully helps:

import seaborn as sns
sns.lineplot(x='x', y='y', hue='node', marker='o', data=df)

Answer 2

Found a solution thankts to some tipps in the comments.

Now i use matplotlip to plot and pandas to group:

fig = plt.figure(num=None, figsize=(15, 15), dpi=80, facecolor='w', edgecolor='k')
ax = fig.subplots()

df = pd.read_csv('Filename', delimiter=' ', names=["node", "time", "x","y"], header=None)

# Scatter for every node
for node, node_data in df.groupby('node'):
    ax.scatter(node_data['x'], node_data['y'], label='node ' + str(node), s=2,
    c=np.random.rand(3,))
    #or use ax.plot to show lines

plt.show()