Plotting basic raman spectra

[sabourizk]

Hi, I'm new to spectrochempy and would love some help plotting my raman data. There are there different lines (defined in my y dimension). I'd like to plot them, create a legend with labels I create (and not the self-generated ones) and axis limits, and exclude any data points that exceed a certain value. Any help would be very much appreciated. I'm not sure why I keep getting stuck.

I tried using the plot_multiple method, but I want all the lines to be solid (and not dashed or dotted). I tried just using the data.plot() method but I can't figure out how to remove data points above a certain value in my y axis (or my "data" dimension).

Thanks!

[fernandezc]

For plot multiple you can use the new version of this function (not yet released):
https://www.spectrochempy.fr/latest/reference/generated/spectrochempy.NDDataset.html#spectrochempy.NDDataset.plot_multiple

To use it you need to use conda or mamba and install the dev version

mamba install -c spectrocat/label/dev spectrochempy

If you send me some sample spectra, I can write a small exemple to show you how to do the plotting you want (this could be introduced in our documentation as well). To do advanced plotting, it may be sometimes necessary to use directly the matplotlib library.

[fernandezc]

", and exclude any data points that exceed a certain value".

For this you could use masks.

In a next post I will give you an example of code.

[fernandezc]

example of code:

import spectrochempy as scp
A = scp.read_labspec("ramandata/SMC1-Initial_RT.txt")
_ = A.plot(ylim=(0,3500))
A[A.data>2000] = scp.MASKED
_ = A.plot(ylim=(0,3500))

Which gives this:

[sabourizk]

lasertest1.txt

thank you!! I'll keep trying it out. I attached a sample txt file that I've been trying to work with. Thanks so much!

[fernandezc]

Well, I found a bug trying to write a simple example with plot_multiple.
I will try to correct it ASAP.
So don't loose time until a new dev version is done.

[fernandezc]

OK plot_multiple is now fixed. You can get the last DEV version, and here is a sample code on how you can proceed.
(I did not try with your data, but it is probable similar)

import spectrochempy as scp

B1 = scp.read("ramandata/serie190214-1.txt")

# Basic plot
_ = B1.plot(cmap=None, lw=1)

# limit the x range
B2 = B1[:, 60.0:] # note the float number to specify that we use coordinates

# detrend
B3 = scp.detrend(B2)
_ = B3.plot(cmap=None)

# denoising
pca = scp.PCA(n_components=10)
pca.fit(B3)
B4 = pca.inverse_transform()
_ = B4.plot(cmap=None)

# take only some spectra (5) 
B5 = B4[:5]
B5.plot(cmap=None)

# plot multiple with offset and labels
prefs = B5.preferences
prefs.figure.figsize = (8,6) 
datasets = list(B5)
_ = scp.plot_multiple(datasets, 
                      method="pen",
                      legend="best",
                      labels=["A", "B", "C", "D", "E"],
                      color = ["black", "red", "green", "blue", "violet"],
                      lw=[1, 2.5, 1, 1, 1], # line width
                      ls='-',  # solid line
                      shift=1000, # vertical shift
                     )

B1 plot

B3 plot

B4 plot

plot_multiple

[sabourizk]

Thank you!! For some reason, I get this error: | TypeError: spectrochempy.core.dataset.arraymixins.ndplot.NDPlot.plot() got multiple values for keyword argument 'ls' for B5. I tried linestyle = '-' and that worked though! thanks so much!

[fernandezc]

I open an issue from this discussion in order to not forget to look at this problem.

[fernandezc]

In principle this is now solved : see #690 (comment).

I mark this discussion as answered for now.