matplotlib.pyplot shows "old data"

[sschusser]

What happened?

If I try to plot "plain data" in pyplot (plt), the plot shows "old data" from spectrochempy NDDataset, too.

Maybe it is related to inline plot configuration in Jupyter? I do not configure any special plotting options but spectrochempy seems to do so automatically when imported or plotting methode is called.

The following picture shows the two plots I get. Surprisingly, they show the same contend and share the same consecutive figure number.

What did you expect to happen?

Since I configured two independent plots, I expect two seperated plots as output.

Minimal Complete Verifiable Example

import spectrochempy as scp
from matplotlib import pyplot as plt

A = scp.NDDataset.linspace(1,100,50)
A = np.sqrt(A)
ax1 = A.plot()
plt.show()

plt.plot([0,10,20],[2,4,7])
plt.show()

Relevant log output

No response

Anything else we need to know?

No response

Environment

INSTALLED VERSIONS

commit: None
python: 3.10.9 (tags/v3.10.9:1dd9be6, Dec 6 2022, 20:01:21) [MSC v.1934 64 bit (AMD64)]
python-bits: 64
OS: Windows
OS-release: 10
machine: AMD64
processor: Intel64 Family 6 Model 140 Stepping 1, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: ('en_US', 'ISO8859-1')

[fernandezc]

Actually, this is not a SpectroChemPy bug, but a matplotlib feature!

To demonstrate I modified your example to use only matplotlib

import numpy as np
from matplotlib import pyplot as plt

fig = plt.figure()
A = np.sqrt(np.linspace(1,100,50))
plt.plot(A)
plt.plot([0,10,20],[2,4,7])
plt.show()

here you get a single figure with the two plot together: this is expected

import numpy as np
from matplotlib import pyplot as plt

fig = plt.figure()
A = np.sqrt(np.linspace(1,100,50))
plt.plot(A)
plt.show()
plt.plot([0,10,20],[2,4,7])
plt.show()

Here you get two time the same plot.

import numpy as np
from matplotlib import pyplot as plt

fig = plt.figure()
A = np.sqrt(np.linspace(1,100,50))
plt.plot(A)
plt.show()

fig2 = plt.figure()
plt.plot([0,10,20],[2,4,7])
plt.show()

Now you get to separate plot. So do not forget to instantiate a new figure for the second separate plot.

SpectroChempy instantiates a new figure each time you use the plot method of a NDDataset. To have two plot on single figure you use clear=False option and specified the ax of the previous plot.

Hope it helps.

[sschusser]

That means there is now chance to use the implicite style of pyplot for quick drawings?

[fernandezc]

Not completely sure of what you mean about quick drawings, but I will try to answer.

First instead of discussing on a closed issue, I will convert it to an open discussion, as the answers may be useful to other users.

[fernandezc]

SpectroChemPy uses matplotlib's explicit style by default (see https://www.spectrochempy.fr/latest/userguide/plotting/plotting.html for a detailed description - where unfortunately I've just seen that there are some warning problems - I'll look into this later).
In fact, as the plot method returns a matplotlib Axe object, you can easily modify it using appropriate methods.

So we can write:

With arguments to the plot method

import spectrochempy as scp
A = scp.NDDataset(....something...)
ax = A.plot(xlim=(40,50))
scp.show()

or with explicit style

import spectrochempy as scp
A = scp.NDDataset(....something...)
ax = A.plot()
ax.set_xlim(40,50)
scp.show()

or with the implicit style (which requires importing matplotlib.pyplot again, but this is done without any particular cost since scp has already imported the library)

import spectrochempy as scp
A = scp.NDDataset(....something...)
ax = A.plot()

import matplolib.pyplot as plt 
plt.xlim(40,50)
plt.show()

You can also rewrite your own plot function using A data attribute.