# -*- coding: utf-8 -*-
# written by Ralf Biehl at the Forschungszentrum Jülich ,
# Jülich Center for Neutron Science 1 and Institute of Complex Systems 1
# Jscatter is a program to read, analyse and plot data
# Copyright (C) 2015-2019 Ralf Biehl
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
#
"""
This is a rudimentary interface to `matplotlib <https://matplotlib.org/>`_ to use dataArrays/sasImage easier.
The standard way to use matplotlib is full available without using this module.
Nevertheless the source can be used as template to be adapted.
You may switch to use mpl in fitting and examples using ::
js.usempl(True)
The intention is to allow fast/easy plotting (one command to plot) with some convenience
function in relation to dataArrays and in a non blocking mode of matplotlib.
E.g. to include automatically the value of an attribute (qq in example) in the legend::
fig[0].Plot(mydataArray, legend='sqr=$qq',sy=[2,3,-1],li=0)
# dataList
fig[0].Plot(mydataList , legend='sqr=$qq',sy=[2,3,-1],li=0)
With somehow shorter form to determine the marker (sy=symbol) and line (li)
and allow plotting in one line. Matplotlib is quite slow (and looks for me ugly).
For 2D plotting use xmgrace.
For 3D plotting this will give some simple plot options (planned).
* The new methods introduced all start with a big Letter to allow still the access of the original methods.
* By indexing subplots can be accessed as figure[i] which is figure.axes[i].
* Same for axes with lines figure[0][i] is figure.axes[0].lines[i].
Example 1::
import jscatter as js
import numpy as np
i5=js.dL(js.examples.datapath+'/iqt_1hho.dat')
p=js.mplot()
p[0].Plot(i5,sy=[-1,0.4,-1],li=1,legend='Q= $q')
p[0].Yaxis(scale='l')
p[0].Title('intermediate scattering function')
p[0].Legend(x=1.13,y=1) # x,y in relative units of the plot
p[0].Yaxis(label='I(Q,t)/I(Q,0)',min=0.01)
p[0].Xaxis(label='Q / 1/nm',max=120)
Example 2 ( same as js.mpl.test() )::
import jscatter as js
import numpy as np
from matplotlib import pyplot
# use this
#fig=pyplot.figure(FigureClass=js.mpl.Figure)
# or
fig=js.mplot()
fig.Multi(2,1)
fig[0].SetView(0.1,0.25,0.8,0.9)
fig[1].SetView(0.1,0.09,0.8,0.23)
q=js.loglist(0.01,5,100)
aa=js.dL()
for pp in range(5):
aa.append(js.dA(np.c_[q,-pp*np.sin(q),0.2*np.cos(5*q)].T))
aa[-1].qq=pp
bb=js.dA(np.c_[q,q**2].T)
bb.qq=123
for pp in range(5):
fig[0].Plot(aa[pp].X,-1*aa[pp].Y,legend='some stufff',sy=[1,(pp+1)/10.],li=0)
fig[0].Plot(aa, legend='qq = $qq', sy=[-1, 0.4, -1, ''], li=0, markeredgewidth=1)
for pp in range(5):
fig[1].Plot(aa[-1].X/5+pp,pp*aa[-1].Y,legend='q=%.1f' %pp,sy=0,li=-1,markeredgewidth =1)
fig[1].Plot(bb,legend='sqr=$qq ',sy=2,li=2)
fig[0].Title('test')
fig[0].Legend(x=1.3,y=1)
fig[1].Legend(x=1.3,y=1)
fig[0].Yaxis(label='y-axis')
fig[1].Yaxis(label='something else')
fig[0].tick_params(labelbottom=False)
fig[1].Xaxis(label='x-axis')
**Some short hints for matplotlib**
Dont use the pyplot interface as it hides how most things work and e.g. how to access lines later.
See `THIS <http://pbpython.com/effective-matplotlib.html>`_ .
After fitting the errorplot can be accessed as ``data.errplot``.
::
fig=js.mplot() # access figure properties from fig
fig.axes[0] # access to axes properties
fig.axes[0].lines[0] # access to lines properties in axes 0
fig.axes[0].lines[1].set_color('b') # change color
fig.axes[0].legend(...) # set legend
data.errplot.axes[0].set_yscale('log') # set log scale in errplot
# for more read matplotlib documentation
"""
from functools import reduce
import copy
import numbers
import matplotlib
# this import of Axes3D is needed to use projection='3d' allover in jscatter
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.projections import register_projection
from matplotlib import pyplot
from matplotlib.lines import Line2D
from matplotlib import colors
import numpy as np
# Use headless mode as general option if no X-display is present
if matplotlib.get_backend() in matplotlib.rcsetup.non_interactive_bk:
_headless = True
else:
_headless = False
lineStyles = ('', '-', '--', '-.', ':')
# linecolors = ('w', 'k', 'r', 'b', 'g', 'c', 'm', 'y',)
linecolors = ('white', 'black', 'red', 'darkgreen', 'blue', 'grey', 'orange', 'magenta', 'yellow', 'green')
fillstyles = ('none', 'full', 'left', 'right', 'bottom', 'top',)
symboldefault = [1, 0.3, 1, ''] # type,size,facecolor,edgecolor
linedefault = [1, 0.5, 1] # type,size,color
#: gracefactor to get same scaling as in grace set to 10
gf = 10
def _translate(axlen, kwargs, data=None, yerr=None):
"""
This function transforms a short description as [1,2,3] for symbol and line to matplotlib compatible arguments.
This allows a shorter description of the symbol and line formats.
Additionally the replacement of $parname in dataArray attributes is done.
"""
# split some special keywords in kwargs
if 'legend' in kwargs:
legend = kwargs['legend']
del kwargs['legend']
elif 'le' in kwargs:
legend = kwargs['le']
del kwargs['le']
else:
legend = None
if 'line' in kwargs:
line = kwargs['line']
del kwargs['line']
elif 'li' in kwargs:
line = kwargs['li']
del kwargs['li']
else:
line = ''
if 'symbol' in kwargs:
symbol = kwargs['symbol']
del kwargs['symbol']
elif 'sy' in kwargs:
symbol = kwargs['sy']
del kwargs['sy']
else:
symbol = [-1, 0.3, -1]
if 'errorbar' in kwargs:
errorbar = kwargs['errorbar']
del kwargs['errorbar']
elif 'er' in kwargs:
errorbar = kwargs['er']
del kwargs['er']
else:
errorbar = None
# replace $attr by the value in data
if legend is not None:
if '$' in legend and hasattr(data, '_isdataArray'):
for par in data.attr:
if '$' + par in legend or '$(' + par + ')' in legend:
# noinspection PyBroadException
try:
vall = np.array(getattr(data, par)).flatten()[0]
if isinstance(vall, numbers.Number):
val = '%.4g' % vall
else:
val = str(vall)
if '$(' + par + ')' in legend:
legend = legend.replace('$(' + par + ')', val)
else:
legend = legend.replace('$' + par, val)
except:
pass
# --------
if isinstance(symbol, (int, str)):
symbol = [symbol] # type,size,facecolor,edgecolor
symbol += symboldefault[len(symbol):]
# symbol marker
if isinstance(symbol[0], numbers.Number):
if symbol[0] < 0: symbol[0] = axlen
if symbol[0] > 0:
symbol[0] = Line2D.filled_markers[divmod(symbol[0] - 1, len(Line2D.filled_markers))[1]]
else:
symbol[0] = None
# symbol color
if isinstance(symbol[2], numbers.Number):
if symbol[2] < 0: symbol[2] = axlen
if symbol[2] > 0:
symbol[2] = linecolors[divmod(symbol[2] - 1, len(linecolors) - 1)[1] + 1]
else:
symbol[2] = None
# edgecolor
if isinstance(symbol[3], numbers.Number):
if symbol[3] < 0: symbol[3] = axlen
if symbol[3] > 0:
symbol[3] = linecolors[divmod(symbol[3] - 1, len(linecolors) - 1)[1] + 1]
else:
symbol[3] = linecolors[0]
else:
# synchronize with facecolor
symbol[3] = symbol[2]
# same for line
if isinstance(line, (int, str)):
# type,size,color
line = linedefault[:2] + [line]
if isinstance(line[0], numbers.Number): # type
if line[0] < 0: line[0] = axlen
if line[0] > 0:
line[0] = lineStyles[divmod(line[0] - 1, len(lineStyles) - 1)[1] + 1]
else:
line[0] = None
if isinstance(line[2], numbers.Number): # color
if line[2] < 0: line[2] = axlen
if line[2] > 0:
line[2] = linecolors[divmod(line[2] - 1, len(linecolors) - 1)[1] + 1]
else:
line[0] = '' # this makes no line to overwrite default '-'
line[2] = None
if symbol[0] is None and line[2] is not None:
symbol[2] = line[2]
if yerr is None or errorbar is None:
errorbar = [None, None]
else:
if isinstance(errorbar, numbers.Number):
errorbar = [None, errorbar]
if isinstance(errorbar[0], numbers.Number):
if errorbar[0] < 0: errorbar[0] = axlen
if errorbar[0] > 0:
errorbar[0] = linecolors[divmod(errorbar[0] - 1, len(linecolors) - 1)[1] + 1]
else:
errorbar[0] = linecolors[0]
else:
errorbar[0] = None
# fmt=fmt,markersize=ssize, markerfacecolor=mfc,linewidth=lsize,label=legend
# capsize same as markersize
for opt, val in zip(['color', 'marker', 'linestyle', 'markersize', 'markerfacecolor', 'markeredgecolor',
'linewidth', 'elinewidth', 'ecolor', 'capsize'],
[symbol[2], symbol[0], line[0], symbol[1] * gf, symbol[2], symbol[3],
line[1], errorbar[1], errorbar[0], symbol[1] * gf / 3.]):
if opt not in kwargs:
kwargs[opt] = val
if legend is not None:
if r'\n' in legend:
lines = []
for line in legend.replace('\\n', '\n').splitlines():
words = line.split()
if len(words) > 4:
line = ' '.join(words[:5]) + '...'
else:
line = ' '.join(words)
lines.append(line)
kwargs['label'] = '\n'.join(lines)
else:
kwargs['label'] = legend
return kwargs
# noinspection PyIncorrectDocstring,PyIncorrectDocstring,PyIncorrectDocstring,PyIncorrectDocstring,PyIncorrectDocstring
[docs]class jspaperAxes(matplotlib.axes.Axes):
"""
An Axes that should look like typical paper layout.
"""
name = 'paper'
def __init__(self, *args, **kwargs):
super(matplotlib.axes.Axes, self).__init__(*args, **kwargs)
self.tick_params(axis='both', direction='in')
[docs] def SetView(self, xmin=None, ymin=None, xmax=None, ymax=None):
"""
This sets the bounding box of the axes.
Parameters
----------
xmin,xmax,ymin,ymax : float
view range
"""
self.set_position([xmin, ymin, xmax - xmin, ymax - ymin]) # [left, bottom, width, height]
self.figure.show()
def __getitem__(self, key):
return self.lines[key]
# noinspection PyIncorrectDocstring
[docs] def plot(self, *datasets, **kwargs):
"""
Plot dataArrays/dataList or array in matplotlib axes.
Parameters are passed to matplotlib.axes.Axes.plot
Parameters
----------
datasets : dataArray/dataList or 1D arrays
Datasets to plot.
- Can be several dataArray/dataList (with .X, .Y and .eY) or 1D arrays (a[1,:],b[2,:]), but dont mix it.
- If dataArray/dataList has .eY errors a errorbars are plotted.
- If format strings are found only the first is used. symbol, line override this.
- Only a single line for 1D arrays is allowed.
symbol,sy : int, list of float
- [symbol,size,color,fillcolor,fillpattern] as [1,1,1,-1];
- single integer to chose symbol eg symbol=3; symbol=0 switches off
- negative increments from last
- symbol => see Line2D.filled_markers
- size => size in pixel
- color => int in sequence = wbgrcmyk
- fillcolor=None see color
- fillpattern=None 0 empty, 1 full, ....test it
line,li : int, list of float or Line object
- [linestyle,linewidth,color] as [1,1,''];
- negative increments
- single integer to chose linestyle line=1; line=0 switches of
- linestyle int '-','--','-.',':'
- linewidth float increasing thickness
- color see symbol color
errorbar,er : int or list of float or Errorbar object
- [color,size] as [1,1]; no increment, no repeat
- color int see symbol color, non-integer syncs to symbol color
- size float default 1.0 ; smaller is 0.5
legend,le : string
- determines legend for all datasets
- string replacement: attr name prepended by '$' (eg. '$par')
is replaced by value str(par1.flatten()[0]) if possible.
$(par) for not unique names
errorbar,er : float
- errorbar thickness, zero is no errorbar
"""
# extract format strings
fmt = [dset for dset in datasets if isinstance(dset, str)]
datasets = [dset for dset in datasets if not isinstance(dset, str)]
# concat to dataList's if its not a format string
if np.alltrue([hasattr(dset, '_isdataList') or (hasattr(dset, '_isdataArray') and np.ndim(dset) > 1)
for dset in datasets]):
# use a single list
datasets = reduce(lambda a, b: a + b, datasets)
if hasattr(datasets, '_isdataArray'):
# return always as dataList not only dataArray
datasets = [datasets]
# If 1 dim data are given
elif np.alltrue([np.ndim(dset) == 1 for dset in datasets]):
# We create a single dataset and use this
shape0 = [np.shape(dset)[0] for dset in datasets]
if shape0.count(shape0[0]) == len(shape0):
# all same length -> make array
datasets = [np.asanyarray(datasets)]
else:
raise TypeError('Dont know how to plot this.')
# self.lines is updated only after show so we need to count explicitly
nlines = len(self.lines)
showerr = True
if 'comment' in kwargs: del kwargs['comment']
if 'errorbar' in kwargs:
if not kwargs['errorbar']: showerr = False
elif 'er' in kwargs:
if not kwargs['er']: showerr = False
for data in datasets:
if hasattr(data, '_isdataArray'):
if hasattr(data, '_iey') and showerr:
yerr = data.eY
else:
yerr = None
nkwargs = _translate(nlines + 1, kwargs.copy(), data, yerr)
if fmt and 'fmt' not in nkwargs:
# if fmt not empty and not other setting found
nkwargs['fmt'] = fmt[0]
self.errorbar(x=data.X, y=data.Y, yerr=yerr, **nkwargs)
nlines += 1
elif hasattr(data, '_isdataList'):
for da in data:
if hasattr(da, '_iey') and showerr:
yerr = da.eY
else:
yerr = None
nkwargs = _translate(nlines + 1, kwargs.copy(), da, yerr)
if fmt and 'fmt' not in nkwargs:
# if fmt not empty and not other setting found
nkwargs['fmt'] = fmt[0]
self.errorbar(x=da.X, y=da.Y, yerr=yerr, **nkwargs)
nlines += 1
elif isinstance(data, np.ndarray):
if showerr:
# noinspection PyBroadException
try:
yerr = data[2]
except:
yerr = None
nkwargs = _translate(nlines + 1, kwargs.copy(), data, yerr)
if fmt and 'fmt' not in nkwargs:
# if fmt not empty and not other setting found
nkwargs['fmt'] = fmt[0]
self.errorbar(x=data[0], y=data[1], yerr=yerr, **nkwargs)
nlines += 1
self.Autoscale()
self.figure.show()
Plot = plot
[docs] def Yaxis(self, min=None, max=None, label=None, scale=None, size=None, charsize=None, tick=None, ticklabel=None,
**kwargs):
"""
Set xaxis
Parameters
----------
label : string
Label
scale : 'log', 'normal'
Scale
min,max : float
Set min and max
size : int
Pixelsize of label
"""
# TODO: log scale errplot not working in makeErrPlot while setting it afterwards works
if size is not None:
size *= gf
if label is not None:
self.set_ylabel(label, size=size)
if scale is not None and scale[0] == 'l':
if min is None: min = 0.1
if max is None: max = 10
self.set_ylim(min, max)
if scale is not None:
if scale[0] == 'l':
self.set_yscale(value='log', nonpositive='clip', subs=[2, 3, 4, 5, 6, 7, 8, 9])
else:
self.set_yscale(value='linear')
self.figure.show()
[docs] def Xaxis(self, min=None, max=None, label=None, scale=None, size=None, charsize=None, tick=None, ticklabel=None,
**kwargs):
"""
Set xaxis
Parameters
----------
label : string
Label
scale : 'log', 'normal'
Scale
min,max : float
Set min and max of scale
size : int
Pixelsize of label
"""
if size is not None:
size *= gf
if label is not None:
self.set_xlabel(label, size=size)
if scale is not None and scale[0] == 'l':
if min is None: min = 0.1
if max is None: max = 10
self.set_xlim(min, max)
if scale is not None:
if scale[0] == 'l':
self.set_xscale(value='log', nonpositive='clip', subs=[2, 3, 4, 5, 6, 7, 8, 9])
else:
self.set_xscale(value='linear')
self.figure.show()
[docs] def Resetlast(self, ):
pass
[docs] def Legend(self, **kwargs):
"""
Show/update legend.
Parameters
----------
charsize, fontsize : int, default 12
Font size of labels
labelspacing : int , default =12
Spacing of labels
loc : int [0..10] default 1 'upper right'
Location specifier
- ‘best’ 0, ‘upper right’ 1, ‘upper left’ 2, ‘lower left’ 3, ‘lower right’ 4,‘center left’ 6,
x,y : float [0..1]
Determines, **if both** given, loc and sets position in axes coordinates.
Sets bbox_to_anchor=(x,y). Values outside [0,1] are ignored.
kwargs : kwargs of axes.legend
Any given kwarg overrides the previous
"""
if 'charsize' in kwargs:
kwargs['fontsize'] = kwargs.pop('charsize') * 10.
if 'fontsize' not in kwargs: kwargs['fontsize'] = 10
if 'labelspacing' not in kwargs: kwargs['labelspacing'] = 0.2
if 'loc' not in kwargs: kwargs['loc'] = 0 # best
x = kwargs.pop('x', None)
if x is not None and ((x>1) or (x<0)): x = None
y = kwargs.pop('y', None)
if y is not None and ((y > 1) or (y < 0)): x = None
_ = kwargs.pop('position', None)
if x is not None and y is not None:
kwargs['loc'] = 'upper right'
kwargs['bbox_to_anchor'] = (x, y)
self.legend(**kwargs)
self.figure.show()
[docs] def Title(self, title, size=None, **kwargs):
"""set figure title"""
if size is not None:
kwargs.update({'size': size * gf})
self.figure.suptitle(title, **kwargs)
self.figure.show()
[docs] def Subtitle(self, subtitle, size=None, **kwargs):
"""
Append subtitle to title
"""
if size is not None:
kwargs.update({'size': size * gf})
# subtitle=self.get_title()+'\n'+subtitle
self.set_title(subtitle, **kwargs)
[docs] def Clear(self):
"""
Clear data of this axes.
To clear everything use clear().
"""
while len(self.lines):
_ = self.lines.pop()
self.figure.show()
[docs] def Text(self, string, x, y, **kwargs):
size = kwargs.pop('charsize', None)
rot = kwargs.pop('rot', None)
if size is not None: kwargs.update({'size': size * gf})
color = kwargs.pop('color', None)
if isinstance(color, numbers.Number):
color = linecolors[divmod(color - 1, len(linecolors) - 1)[1] + 1]
if color is not None:
kwargs.update({'color': color})
self.text(x=x, y=y, s=string, **kwargs)
[docs] def linlog(self, *args, **kwargs):
self.semilogx(*args, **kwargs)
[docs] def loglin(self, *args, **kwargs):
self.semilogy(*args, **kwargs)
[docs] def Arrow(self, x1=None, y1=None, x2=None, y2=None, linewidth=None, arrow=None):
"""
Plot an arrow or line.
Parameters
----------
x1,y1,x2,y2 : float
Start/end coordinates in box units [0..1].
linewidth : float
Linewidth
arrow : int or ['-','->','<-','<->']
Type of arrow.
If int it selects from ['-','->','<-','<->']
Returns
-------
"""
if isinstance(arrow, numbers.Integral):
arrow = ['-', '->', '<-', '<->'][arrow]
self.annotate("",
xy=(x1, y1), xycoords='data',
xytext=(x2, y2), textcoords='data',
arrowprops=dict(arrowstyle=arrow, connectionstyle="arc3", linewidth=linewidth))
[docs] def Autoscale(self, **kwargs):
"""
Autoscale, see matplotlib.axes.Axes.autoscale_view() .
"""
self.autoscale(**kwargs)
# register that it can be used as other Axes
register_projection(jspaperAxes)
[docs]def show(**kwargs):
"""
Updates figures or saves figures in noninteractive mode (headless)
In headless mode all figures are save to lastopenedplots{i}.png .
Parameters
----------
kwargs : args
Passed to pyplot.show added by block=False
"""
if _headless:
# save all
for i in pyplot.get_fignums():
pyplot.figure(i).savefig(f'lastopenedplots_{i}.png')
return
kwargs.update(block=False)
pyplot.show(**kwargs)
[docs]def close(*args, **kwargs):
"""
Close figure/s. See matplotlib.pyplot.close .
"""
pyplot.close(*args, **kwargs)
[docs]def mplot(width=None, height=None, **kwargs):
"""
Open matplotlib figure in paper layout with methods to display dataArray/dataList.
Paper layout means white background, black axis.
Plot separates X,Y, eY of dataList automatically.
In interactive mode the figure is shown, in headless these can be saved after plotting.
Parameters
----------
width,height : float
Size of plot in cm.
kwargs :
Keyword args of matplotlib.pyplot.figure .
Returns
-------
matplotlib figure
Notes
-----
- By indexing as the axes subplots can be accessed as figure[i] which is figure.axes[i].
- Same for axes with lines figure[0][i] is figure.axes[0].lines[i].
- Some methods with similar behaviour as in grace are defined (big letter commands)
- matplotlib methods are still available (small letters commands)
"""
inch = 2.54
headless = kwargs.pop('headless', _headless)
if width is not None and height is not None:
kwargs.update({'figsize': (width / inch, height / inch)})
if headless:
pyplot.ioff()
else:
pyplot.ion()
kwargs.update({'FigureClass': jsFigure, 'headless': headless})
fig = pyplot.figure(**kwargs)
return fig
[docs]def regrid(x, y, z, shape=None):
"""
Make a meshgrid from XYZ data columns.
Parameters
----------
x,y,z : array like
Array like data should be quadratic or rectangular.
shape : None, shape or first dimension size
If None the number of unique values in x is used as first dimension.
If integer the second dimension is guessed from size.
Returns
-------
2dim arrays for x,y,z
"""
if shape is None:
shape = len(np.unique(x))
if isinstance(shape, numbers.Number):
shape = (shape, -1)
try:
xx = x.reshape(shape)
except ValueError:
xx = None
try:
yy = y.reshape(shape)
except (ValueError, AttributeError):
yy = None
try:
zz = z.reshape(shape)
except (ValueError, AttributeError):
zz = None
return xx, yy, zz
[docs]def surface(x, y, z, shape=None, levels=8, colorMap='jet', lineMap=None, alpha=0.7):
"""
Surface plot of x,y,z, data
If x,y,z differ because of numerical precision use the shape parameter to give the shape explicitly.
Parameters
----------
x,y,z : array
Data as array
shape : integer, 2x integer
Shape of image with len(x)=shape[0]*shape[1] or only first dimension.
See regrid shape parameter.
levels : integer, array
Levels for contour lines as number of levels or array of specific values.
colorMap : string
Color map name, see showColors.
lineMap : string
Color name for contour lines
b: blue
g: green
r: red
c: cyan
m: magenta
y: yellow
k: black
w: white
alpha : float [0,1], default 0.7
Transparency of surface
Returns
-------
figure
Examples
--------
::
import jscatter as js
import numpy as np
R=8
N=50
qxy=np.mgrid[-R:R:N*1j, -R:R:N*1j].reshape(2,-1).T
qxyz=np.c_[qxy,np.zeros(qxy.shape[0])]
sclattice= js.lattice.scLattice(2.1, 5)
ds=[[20,1,0,0],[5,0,1,0],[5,0,0,1]]
sclattice.rotatehkl2Vector([1,0,0],[0,0,1])
ffs=js.sf.orientedLatticeStructureFactor(qxyz,sclattice,domainsize=ds,rmsd=0.1,hklmax=2)
fig=js.mpl.surface(qxyz[:,0],qxyz[:,1],ffs[3].array)
"""
if np.ndim(x) < 2:
X, Y, Z = regrid(x, y, z, shape)
if X is None or Z is None or Y is None:
raise Exception('x,y,z seem not to be on regular grid.')
cmap = matplotlib.cm.get_cmap(colorMap)
try:
lmap = matplotlib.cm.get_cmap(lineMap)
except ValueError:
lmap = lineMap
fig = figure()
ax = fig.add_subplot(111, projection='3d')
surf = ax.plot_surface(X, Y, Z, cmap=cmap, linewidth=1, antialiased=True, alpha=alpha)
# noinspection PyBroadException
try:
contour = ax.contour3D(X, Y, Z, levels, linewidths=1, cmap=lmap)
except:
contour = ax.contour3D(X, Y, Z, levels, linewidths=1, colors=lmap)
ax.set_xlim([min(x), max(x)])
ax.set_ylim([min(y), max(y)])
ax.set_zlim([min(z), max(z)])
ax.set_xlabel('X axis')
ax.set_ylabel('Y axis')
ax.set_zlabel('Z axis')
fig.colorbar(surf, shrink=0.8) # note that colorbar is a method of the figure, not the axes
fig.tight_layout()
show(block=False)
return fig
[docs]def scatter3d(x, y=None, z=None, pointsize=3, color='k', ax=None):
"""
Scatter plot of x,y,z data points.
Parameters
----------
x,y,z : arrays
Data to plot. If x.shape is Nx3 these points are used.
pointsize : float
Size of points
color : string
Colors for points
ax : axes, default None
Axes to plot inside. If None a new figure is created.
Returns
-------
figure
Examples
--------
::
# ellipsoid with grid build by mgrid
import jscatter as js
import numpy as np
# cubic grid points
ig=js.formel.randomPointsInCube(200)
fig=js.mpl.scatter3d(ig.T)
"""
if np.ndim(x) == 2 and (3 in x.shape):
try:
x, y, z = x.T
except ValueError:
x, y, z = x
if ax is None:
fig = figure()
ax = fig.add_subplot(1, 1, 1, projection='3d')
else:
fig = ax.figure
sc = ax.scatter(x, y, z, s=pointsize, color=color)
mi = np.min([x, y])
ma = np.max([x, y])
ax.set_xlim(mi, ma)
ax.set_ylim(mi, ma)
ax.set_zlim(min(z), max(z))
ax.set_xlabel('X axis')
ax.set_ylabel('Y axis')
ax.set_zlabel('Z axis')
try:
ax.set_aspect("equal")
except NotImplementedError:
# It is not currently possible to manually set the aspect on 3D axes
pass
fig.tight_layout()
# fig.colorbar(scatter ,shrink=0.8) # note that colorbar is a method of the figure, not the axes
show(block=False)
return fig
def _isregularspaced(sequence):
diff = np.diff(np.unique(sequence))
return np.all(np.isclose(diff, diff[0]))
[docs]def contourImage(x, y=None, z=None, levels=None, fontsize=10, colorMap='jet', scale='norm', lineMap=None,
axis=None, origin=None, block=False, invert_yaxis=False, invert_xaxis=False,
linthresh=1, linscale=1, badcolor=None):
"""
Image with contour lines of 3D dataArrays or sasImage/image array.
This is a convenience function to easily plot dataArray/sasImage content and covers not all matplotlib options.
The first pixel is at upper left corner and X is vertical as for images which
is sometimes not intuitive for dataArrays.
Use invert_?axis and origin as needed or adapt the source code to your needs.
Parameters
----------
x,y,z : arrays
x,y,z coordinates for z display in x,y locations.
If x is image_array or sasImage this is used ([0,0] pixel upper left corner).
If x is dataArray we plot like x,y,z=x.X,x.Z,x.Y as dataArray use always .Y as value in X,Z coordinates.
x may be dataArray created from a sasImage using ```image.asdataArray```.
Using .regrid the first .X values is at upper left corner.
levels : int, None, sequence of values
Number of contour lines between min and max or sequence of specific values.
colorMap : string
Get a colormap instance from name.
Standard mpl colormap name (see showColors).
badcolor : float, color
Set the color for bad values (like masked pixel) values in an image.
Default is bad values be transparent.
Color can be matplotlib color as 'k','b' or
float value in interval [0,1] of the chosen colorMap.
0 sets to minimum value, 1 to maximum value.
scale : 'log', 'symlog', default = 'norm'
Scale for intensities.
- 'norm' Linear scale.
- 'log' Logarithmic scale
- 'symlog' Symmetrical logarithmic scale is logarithmic in both the positive
and negative directions from the origin. This works also for only positive data.
Use linthresh, linscale to adjust.
linthresh : float, default = 1
Only used for scale 'sym'.
The range within which the plot is linear (-linthresh to linthresh).
linscale : float, default = 1
Only used for scale 'sym'.
Its value is the number of decades to use for each half of the linear range.
E.g. 10 uses 1 decade.
lineMap : string
Label color
Colormap name as in colorMap, otherwise as cs in in Axes.clabel
* if None, the color of each label matches the color of the corresponding contour
* if one string color, e.g., colors = ‘r’ or colors = ‘red’, all labels will be plotted in this color
* if a tuple of matplotlib color args (string, float, rgb, etc),
different labels will be plotted in different colors in the order specified
fontsize : int, default 10
Size of line labels in pixel
axis : None, 'pixel'
If coordinates should be forced to pixel.
Wavevectors are used only for sasImage using getPixelQ.
invert_yaxis,invert_xaxis : bool
Invert corresponding axis.
origin : 'lower','upper'
Origin of the plot in upper left or lower left corner.
See matplotlib imshow.
block : bool
Open in blocking or non-blocking mode
Returns
-------
figure
Notes
-----
- For irregular distributed points (x,z,y) the point positions can later be added by ::
fig.axes[0].plot(x, y, 'ko', ms=1)
js.mpl.show(block=False)
- dataArray created from sasImage(.asdataArray) need to be complete with out missing pixels.
e.g. using ```image.asdataArray(masked=0)``` or by interpolating the missing pixel.
Otherwise the used matplotlib.tricontour will interpolate which looks different than expected.
Examples
--------
Create log scale image for maskedArray (sasImage). ::
import jscatter as js
import numpy as np
# sets negative values to zero
calibration = js.sas.sasImage(js.examples.datapath+'/calibration.tiff')
fig1=js.mpl.contourImage(calibration)
fig1.suptitle('Calibration lin scale')
fig2=js.mpl.contourImage(calibration,scale='log')
#
# change labels and title
ax=fig2.axes[0]
ax.set_xlabel('qx ')
ax.set_ylabel('qy')
fig2.suptitle('Calibration log scaled')
# in case something is not shown
js.mpl.show(block=False)
Use ``scale='symlog'`` for mixed lin=log scaling to pronounce low scattering. ::
import jscatter as js
import numpy as np
# sets negative values to zero
bsa = js.sas.sasImage(js.examples.datapath+'/BSA11mg.tiff')
fig=js.mpl.contourImage(bsa,scale='sym',linthresh=30, linscale=10)
Other examples ::
import jscatter as js
import numpy as np
# On a regular grid
x,z=np.mgrid[-4:8:0.1,-3:5:0.1]
xyz=js.dA(np.c_[x.flatten(),
z.flatten(),
0.3*np.sin(x*z/np.pi).flatten()+0.01*np.random.randn(len(x.flatten())),
0.01*np.ones_like(x).flatten() ].T)
# set columns where to find X,Y,Z )
xyz.setColumnIndex(ix=0,iy=2,iz=1)
# first X value (here -4) is in [0,0] upper left corner, so we invert the corresponding axis
fig=js.mpl.contourImage(xyz,invert_yaxis=True)
#fig.savefig(js.examples.imagepath+'/contourImage.jpg')
.. image:: ../../examples/images/contourImage.jpg
:align: center
:width: 50 %
:alt: contourImage
If points are missing the tricontour allows interpolation of missing contours.
In this case contour lines are used. ::
# remove each 3rd point that we have missing points
# like random points
x,z=js.formel.randomPointsInCube(1500,0,2).T*10-4
xyz=js.dA(np.c_[x.flatten(),
z.flatten(),
1.3*np.sin(x*z/np.pi).flatten()+0.001*np.random.randn(len(x.flatten()))].T)
xyz.setColumnIndex(ix=0,iy=2,iz=1)
js.mpl.contourImage(xyz)
"""
fig = figure()
ax = fig.add_subplot(1, 1, 1)
# use copy so that we do not mutate the global colormap instance; stupid matplotlib programmers
cmap = copy.copy(matplotlib.cm.get_cmap(colorMap))
if badcolor is not None:
# set bad color
if isinstance(badcolor, numbers.Number):
cmap.set_bad(color=cmap(badcolor))
else:
cmap.set_bad(color=badcolor)
try:
lmap = copy.copy(matplotlib.cm.get_cmap(lineMap))
except ValueError:
lmap = lineMap
# determine the scaling (norm)
# determine vmin,vmax later
if scale[:3] == 'log':
norm = colors.LogNorm(clip=True)
elif scale[:3] == 'sym':
norm = colors.SymLogNorm(clip=True, linthresh=linthresh, linscale=linscale)
else: # default: scale == 'normalize':
norm = colors.Normalize(clip=True)
if np.ndim(x) < 2 or hasattr(x, '_isdataArray'):
if hasattr(x, 'sasImageshape') and np.prod(x.sasImageshape) == x.shape[1]:
# this was created by sasImage.asdataArray with masked pixels set to a value
# because the Ewald sphere is not flat the isregularspaced detection fails
z = x.Y.astype('float')
# treat like image with regular pixels
zz, *rest = regrid(z, None, None, shape=x.sasImageshape)
extend = [np.min(x.Z), np.max(x.Z), np.min(x.X), np.max(x.X)]
norm.autoscale(zz)
im = ax.imshow(zz, cmap=cmap, extent=extend, origin=origin, norm=norm)
if levels is not None:
im.cset = ax.contour(zz, levels=levels, linewidths=1, cmap=lmap,
extent=extend, origin=origin, norm=norm)
im.labels = ax.clabel(im.cset, inline=True, fmt='%1.1f', fontsize=fontsize)
else:
if hasattr(x, '_isdataArray'):
x, y, z = x.X, x.Z, x.Y
z = np.copy(z)
# try to regrid and test if zz regrid worked, otherwise tricontour interpolates
xx, yy, zz = regrid(x, y, z, np.unique(x).shape[0])
if _isregularspaced(x) and _isregularspaced(y) and zz is not None:
extend = [yy[0, 0], yy[0, -1], xx[-1, 0], xx[0, 0]]
norm.autoscale(zz)
im = ax.imshow(zz, cmap=cmap, extent=extend, origin=origin, norm=norm)
if levels is not None:
im.cset = ax.contour(zz, levels=levels, linewidths=1, cmap=lmap, extent=extend, origin=origin,
norm=norm)
im.labels = ax.clabel(im.cset, inline=True, fmt='%1.1f', fontsize=fontsize)
else:
# tricontour plots in wrong orientation so change it
x, y = y, x
extend = [np.min(y), np.max(y), np.min(x), np.max(x)]
norm.autoscale(z)
if isinstance(levels, numbers.Integral):
levels = np.r_[norm.vmin:norm.vmax:levels * 1j]
elif not isinstance(levels, (list, tuple)):
levels = None
# tricontour lines
ax.tricontour(x, y, z, levels=levels, linewidths=1, cmap=lmap, extent=extend, origin=origin, norm=norm)
# tricontour filled
im = ax.tricontourf(x, y, z, levels=levels, cmap=cmap, extent=extend, origin=origin, norm=norm)
fig.colorbar(im) # note that colorbar is a method of the figure, not the axes
else:
# image array, copy protects original from being modified
# we need to take care if it is array or masked_array to copy
# using e.g. norm='log' mask zero values
# sis part is used from sasImage.show
if np.ma.is_masked(x):
# copy including mask
z = np.ma.copy(x)
else:
z = np.copy(x)
if np.issubdtype(z.dtype, np.integer):
# dtype int32 throws sometimes error "TypeError: Cannot cast array data from dtype('int32') ..."
z = z.astype('float')
if axis != 'pixel':
# if it is an sasImage we get xy from getPixelQ
x, y = x.pQaxes()
extend = [np.min(y), np.max(y), np.min(x), np.max(x)]
else:
extend = None
# determine vmax and vmin
norm.autoscale(z)
im = ax.imshow(z, cmap=cmap, extent=extend, origin=origin, norm=norm)
if levels is not None:
im.cset = ax.contour(z, levels=levels, linewidths=1, cmap=lmap, extent=extend, origin=origin, norm=norm)
im.labels = ax.clabel(im.cset, inline=True, fmt='%1.1f', fontsize=fontsize)
fig.colorbar(im) # note that colorbar is a method of the figure, not the axes
if invert_yaxis: ax.invert_yaxis()
if invert_xaxis: ax.invert_xaxis()
show(block=block)
return fig
[docs]def contourOnCube(xy, yz=None, xz=None, shape=None, offset=None, levels=None, colorMap='jet', scale='norm',
block=False, linthresh=1, linscale=1, badcolor=None, ax=None):
"""
Plot 3 2d contourf planes on surface of a cube.
Intended to show 3D perpendicular scattering planes together.
Parameters
----------
xy,yz,xz : array 3xNM
2D data [x,y,z] with shape N*M = NM.
Each is ploted parallel to the plane mentioned in name.
regrid is used to reshape to dimension 3xNxM
shape : list 2x float
2D shape of the above arrays
offset : list 3x float, default 0,0,0
Position of the xy,yz,xz planes in a 3D plot.
levels : int, None, sequence of values
Number of contour lines between min and max or sequence of specific values.
colorMap : string
Get a colormap instance from name.
Standard mpl colormap name (see showColors).
badcolor : float, color
Set the color for bad values (like masked pixel) values in an image.
Default is bad values be transparent.
Color can be matplotlib color as 'k','b' or
float value in interval [0,1] of the chosen colorMap.
0 sets to minimum value, 1 to maximum value.
scale : 'log', 'symlog', default = 'norm'
Scale for intensities.
- 'norm' Linear scale.
- 'log' Logarithmic scale
- 'symlog' Symmetrical logarithmic scale is logarithmic in both the positive
and negative directions from the origin. This works also for only positive data.
Use linthresh, linscale to adjust.
linthresh : float, default = 1
Only used for scale 'sym'.
The range within which the plot is linear (-linthresh to linthresh).
linscale : float, default = 1
Only used for scale 'sym'.
Its value is the number of decades to use for each half of the linear range.
E.g. 10 uses 1 decade.
block : bool
Open in blocking or non-blocking mode
Returns
-------
axes
Examples
--------
::
import jscatter as js
import numpy as np
# detector planes; a real flat detector has z>0
q = np.mgrid[-9:9:51j, -9:9:51j].reshape(2,-1).T
grid= js.sf.scLattice(10/20,20).XYZ
fa = js.cloudscattering.fa_cuboid(*grid[:,:3].T,0.2,0.4,2)
rod0=np.array([[0,0,0,1,0,0]])
qz=np.c_[q,np.zeros_like(q[:,0])] # for z=0
qy=np.c_[q[:,:1],np.zeros_like(q[:,0]),q[:,1:]] # for z=0
qx=np.c_[np.zeros_like(q[:,0]),q] # for z=0
ffz1 = js.ff.orientedCloudScattering3Dff(qz,cloud=rod0, formfactoramp=fa)
ffy1 = js.ff.orientedCloudScattering3Dff(qy,cloud=rod0, formfactoramp=fa)
ffx1 = js.ff.orientedCloudScattering3Dff(qx,cloud=rod0, formfactoramp=fa)
# show as cube surfaces
ax=js.mpl.contourOnCube(ffz1[[0,1,3]].array,ffx1[[1,2,3]].array,ffy1[[0,2,3]].array,offset=[-9,-9,9])
#ax.figure.savefig(js.examples.imagepath+'/contourOnCube.jpg')
.. image:: ../../examples/images/contourOnCube.jpg
:width: 50 %
:align: center
:alt: filledSphere
"""
if ax is None:
fig = figure()
ax = fig.add_subplot(1, 1, 1, projection='3d')
else:
fig = ax.figure
# use copy so that we do not mutate the global colormap instance; stupid matplotlib programmers
cmap = copy.copy(matplotlib.cm.get_cmap(colorMap))
if badcolor is not None:
# set bad color
if isinstance(badcolor, numbers.Number):
cmap.set_bad(color=cmap(badcolor))
else:
cmap.set_bad(color=badcolor)
# try:
# lmap = copy.copy(matplotlib.cm.get_cmap(lineMap))
# except ValueError:
# lmap = lineMap
# determine the scaling (norm)
# determine vmin,vmax later
if scale[:3] == 'log':
norm = colors.LogNorm(clip=True)
elif scale[:3] == 'sym':
norm = colors.SymLogNorm(clip=True, linthresh=linthresh, linscale=linscale)
else: # default: scale == 'normalize':
norm = colors.Normalize(clip=True)
if offset is None:
offset = 0
if isinstance(offset, numbers.Number):
offset=[offset]*3
minmax=[]
for xyz, zdir, off, o in zip([xy, yz, xz], ['z', 'x', 'y'], offset, [[0, 1, 2], [2, 0, 1], [0, 2, 1]]):
if xyz is None:
continue
minmax.append([np.min(xyz[0]), np.max(xyz[0])])
minmax.append([np.min(xyz[1]), np.max(xyz[1])])
XZY = regrid(xyz[0], xyz[1], xyz[2], shape)
if XZY[0] is None or XZY[1] is None or XZY[2] is None:
raise Exception('Some input seems not to be on regular grid.')
# determine vmax and vmin
norm.autoscale(XZY[2])
ax.contourf(XZY[o[0]], XZY[o[1]], XZY[o[2]],
zdir=zdir, offset=off, cmap=colorMap, levels=levels, norm=norm)
minmax = np.array(minmax)
ax.tick_params(labelsize=8, pad=4)
ax.set_xlabel('X axis', fontsize='smaller')
ax.set_ylabel('Y axis', fontsize='smaller')
ax.set_zlabel('Z axis', fontsize='smaller')
ax.set_xlim(minmax[:, 0].min(), minmax[:, 1].max())
ax.set_ylim(minmax[:, 0].min(), minmax[:, 1].max())
ax.set_zlim(minmax[:, 0].min(), minmax[:, 1].max())
# fig.colorbar(surf, shrink=0.8) # note that colorbar is a method of the figure, not the axes
# fig.tight_layout()
show(block=block)
return ax
[docs]def showColors():
"""
Get a list of the colormaps in matplotlib.
Ignore the ones that end with '_r' because these are
simply reversed versions of ones that don't end with '_r'
Colormaps Names
Accent, Accent_r, Blues, Blues_r, BrBG, BrBG_r, BuGn, BuGn_r, BuPu, BuPu_r,
CMRmap, CMRmap_r, Dark2, Dark2_r, GnBu, GnBu_r, Greens, Greens_r, Greys,
Greys_r, OrRd, OrRd_r, Oranges, Oranges_r, PRGn, PRGn_r, Paired, Paired_r,
Pastel1, Pastel1_r, Pastel2, Pastel2_r, PiYG, PiYG_r, PuBu, PuBuGn,
PuBuGn_r, PuBu_r, PuOr, PuOr_r, PuRd, PuRd_r, Purples, Purples_r,
RdBu, RdBu_r, RdGy, RdGy_r, RdPu, RdPu_r, RdYlBu, RdYlBu_r, RdYlGn,
RdYlGn_r, Reds, Reds_r, Set1, Set1_r, Set2, Set2_r, Set3, Set3_r,
Spectral, Spectral_r, Vega10, Vega10_r, Vega20, Vega20_r, Vega20b,
Vega20b_r, Vega20c, Vega20c_r, Wistia, Wistia_r, YlGn, YlGnBu, YlGnBu_r,
YlGn_r, YlOrBr, YlOrBr_r, YlOrRd, YlOrRd_r, afmhot, afmhot_r, autumn,
autumn_r, binary, binary_r, bone, bone_r, brg, brg_r, bwr, bwr_r, cool,
cool_r, coolwarm, coolwarm_r, copper, copper_r, cubehelix, cubehelix_r,
flag, flag_r, gist_earth, gist_earth_r, gist_gray, gist_gray_r, gist_heat,
gist_heat_r, gist_ncar, gist_ncar_r, gist_rainbow, gist_rainbow_r, gist_stern,
gist_stern_r, gist_yarg, gist_yarg_r, gnuplot, gnuplot2, gnuplot2_r, gnuplot_r,
gray, gray_r, hot, hot_r, hsv, hsv_r, inferno, inferno_r, jet, jet_r, magma,
magma_r, nipy_spectral, nipy_spectral_r, ocean, ocean_r, pink, pink_r, plasma,
plasma_r, prism, prism_r, rainbow, rainbow_r, seismic, seismic_r, spectral,
spectral_r, spring, spring_r, summer, summer_r, tab10, tab10_r, tab20, tab20_r,
tab20b, tab20b_r, tab20c, tab20c_r, terrain, terrain_r, viridis, viridis_r,
winter, winter_r
From
https://matplotlib.org/1.2.1/examples/pylab_examples/show_colormaps.html
"""
a = np.linspace(0, 1, 256).reshape(1, -1)
a = np.vstack((a, a))
# Get a list of the colormaps in matplotlib. Ignore the ones that end with
# '_r' because these are simply reversed versions of ones that don't end
# with '_r'
maps = sorted(m for m in matplotlib.cm.cmap_d if not m.endswith("_r"))
nmaps = len(maps) + 1
#
fig = figure(figsize=(5, 10))
fig.subplots_adjust(top=0.99, bottom=0.01, left=0.2, right=0.99)
for i, m in enumerate(maps):
ax = fig.add_subplot(nmaps, 1, i + 1)
ax.axis("off")
ax.imshow(a, aspect='auto', cmap=matplotlib.cm.get_cmap(m), origin='lower')
pos = list(ax.get_position().bounds)
fig.text(pos[0] - 0.01, pos[1], m, fontsize=10, horizontalalignment='right')
#
show(block=False)
return fig
[docs]def showlastErrPlot2D(data, lastfit=None, shape=None, scale='norm', colorMap='jet', method='nearest',
linthresh=1, linscale=1, badcolor=None, transpose=None, figsize=[6, 6],
txtkwargs={'fontsize': 'small'}):
"""
Show a 2D errplot for 2D fit data.
Parameters
----------
data : dataArray
dataArray optional with fit values in lastfit.
lastfit : None, dataArray
Lastfit dataArray if not present in data.
Can be used to create showlastErrPlot2D from saved data and lastfit.
shape : [int,int]
Optional shape of the data if these are from an image.
If not given the data are interpolated (regrid)
method : float,'linear', 'nearest', 'cubic'
Filling value for new points as float or order of interpolation
between existing points.
See `griddata <https://docs.scipy.org/doc/scipy/reference/generated/scipy.interpolate.griddata.html>`_
colorMap : string
Get a colormap instance from name.
Standard mpl colormap name (see showColors).
badcolor : float, color
Set the color for bad values (like masked pixel) values in an image.
Default is bad values be transparent.
Color can be matplotlib color as 'k','b' or
float value in interval [0,1] of the chosen colorMap.
0 sets to minimum value, 1 to maximum value.
scale : 'log', 'symlog', default = 'norm'
Scale for intensities.
- 'norm' Linear scale.
- 'log' Logarithmic scale
- 'symlog' Symmetrical logarithmic scale is logarithmic in both the positive
and negative directions from the origin. This works also for only positive data.
Use linthresh, linscale to adjust.
linthresh : float, default = 1
Only used for scale 'sym'.
The range within which the plot is linear (-linthresh to linthresh).
linscale : float, default = 1
Only used for scale 'sym'.
Its value is the number of decades to use for each half of the linear range.
E.g. 10 uses 1 decade.
transpose : bool
Transpose coordinates, e.g. for sasImages.
figsize : [float,float], default [6,6]
Figure Size in inch.
txtkwargs : kwargs
Keyword arguments passed to Text https://matplotlib.org/api/text_api.html#matplotlib.text.Text).
except x,y,text arguments.
Examples
--------
::
import jscatter as js
import numpy as np
# create 2D data with X,Z axes and Y values as Y=f(X,Z)
x,z=np.mgrid[-5:3:0.05,-5:9:0.05]
xyz=js.dA(np.c_[x.flatten(),
z.flatten(),
0.3*np.sin(x*z/np.pi).flatten()+0.01*np.random.randn(len(x.flatten())),
0.01*np.ones_like(x).flatten() ].T)
# set columns where to find X,Y,Z )
xyz.setColumnIndex(ix=0,iz=1,iy=2,iey=3)
#
ff=lambda x,z,a,b:a*np.sin(b*x*z)
xyz.fit(ff,{'a':1,'b':1/3.},{},{'x':'X','z':'Z'})
fig = js.mpl.showlastErrPlot2D(xyz)
#fig.savefig(js.examples.imagepath+'/2dfitgoodfit2.jpg')
xyz.save('dat.dat') # save data
xyz.lastfit.save('lastfit.dat') # save lastfit
::
# recover from saved data above
fig = js.mpl.showlastErrPlot2D(js.dA('dat.dat'),js.dA('lastfit.dat'))
.. image:: ../../examples/images/2dfitgoodfit2.jpg
:align: center
:width: 70 %
:alt: 2dfitgoodfit2
::
import jscatter as js
import numpy as np
import matplotlib.pyplot as pyplot
import matplotlib.tri as tri
randn=np.random.randn
rand=np.random.rand
def somepeaks(width, height,a,b,c):
return a*width*(1-width)*np.cos(b*np.pi*width) * np.sin(c*np.pi*height**2)**2
# create random points in [0,1]
NN=1000
xz = rand(NN, 2)
v = somepeaks(xz[:,0], xz[:,1],1,4,4)
# create dataArray
data=js.dA(np.stack([xz[:,0], xz[:,1],v+0.01*randn(NN),np.ones(NN)*0.01]), XYeYeX=[0, 2, 3, None, 1, None])
# bad start parameters
data.fit(somepeaks,{'a':1,'b':2,'c':1},{},{'width':'X','height':'Z'})
fig = js.mpl.showlastErrPlot2D(data)
# good start parameters
data.fit(somepeaks,{'a':0.8,'b':3.8,'c':4.2},{},{'width':'X','height':'Z'})
fig = js.mpl.showlastErrPlot2D(data)
#fig.savefig(js.examples.imagepath+'/2dfitgoodfit.jpg')
.. image:: ../../examples/images/2dfitgoodfit.jpg
:align: center
:width: 70 %
:alt: 2dfitgoodfit
"""
# use copy so that we do not mutate the global colormap instance; stupid matplotlib programmers
cmap = copy.copy(matplotlib.cm.get_cmap(colorMap))
if hasattr(data, 'W') and shape is None:
print('This function does not yet handle 3D data!')
return
if badcolor is not None:
# set bad color
if isinstance(badcolor, numbers.Number):
cmap.set_bad(color=cmap(badcolor))
else:
cmap.set_bad(color=badcolor)
# determine the scaling (norm)
# determine vmin,vmax later
if scale[:3] == 'log':
norm1 = colors.LogNorm(clip=True)
elif scale[:3] == 'sym':
norm1 = colors.SymLogNorm(clip=True, linthresh=linthresh, linscale=linscale)
else: # default: scale == 'normalize':
norm1 = colors.Normalize(clip=True)
norm2 = colors.Normalize(clip=True)
if lastfit is None:
lastfit = data.lastfit
else:
lastfit.getfromcomment('func_name')
lastfit.getfromcomment('func_code')
if shape is not None:
X, Z, Y = regrid(data.X, data.Z, data.Y, shape)
lf_Y, _, _ = regrid(lastfit.Y, None, None, shape)
extend = [np.min(Z), np.max(Z), np.min(X), np.max(X)]
elif _isregularspaced(data.X) and _isregularspaced(data.Z):
X, Z, Y = regrid(data.X, data.Z, data.Y, np.unique(data.X).shape[0])
lf_Y, _, _ = regrid(lastfit.Y, None, None, np.unique(data.X).shape[0])
extend = [np.min(Z), np.max(Z), np.min(X), np.max(X)]
else:
extend = [np.min(data.Z), np.max(data.Z), np.min(data.X), np.max(data.X)]
dx = extend[1] - extend[0]
dy = extend[3] - extend[2]
nn = data.shape[1] ** 0.5 * 2
nnx = int(nn * dx / dy)
nnz = int(nn * dy / dx)
newdata = data.regrid(nnx, nnz, 1, method=method, fill_value=0)
Y, _, _ = regrid(newdata.Y, None, None, nnx)
lf_Y, _, _ = regrid(lastfit.regrid(nnx, nnz, 1, method=method, fill_value=0).Y, None, None, nnx)
if transpose:
Y = Y.T
lf_Y = lf_Y.T
extend = [extend[2], extend[3], extend[0], extend[1]]
fig = figure(figsize=figsize)
ax1 = fig.add_subplot(2, 2, 1)
ax2 = fig.add_subplot(2, 2, 2)
ax3 = fig.add_subplot(2, 2, 3)
dY = lf_Y - Y
# add images and and some text
norm1.autoscale(Y)
im1 = ax1.imshow(Y, cmap=cmap, norm=norm1, extent=extend, origin='lower')
im2 = ax2.imshow(lf_Y, cmap=cmap, norm=norm1, extent=extend, origin='lower')
norm2.autoscale(dY)
im3 = ax3.imshow(dY, cmap=cmap, norm=norm2, extent=extend, origin='lower')
# fig.colorbar(im1, ax=ax1, pad=0.01, shrink=0.85)
c2 = fig.colorbar(im2, ax=[ax1, ax2], pad=0.01, shrink=0.85)
c3 = fig.colorbar(im3, ax=ax3, pad=0.01, shrink=0.85)
ax1.set_ylabel('Z')
ax1.set_xlabel('X')
# ax2.set_ylabel('Z')
ax2.set_xlabel('X')
ax3.set_ylabel('Z')
ax3.set_xlabel('X')
ax2.set_yticklabels([])
ax1.set_title('original')
ax2.set_title('fit')
ax3.set_title('difference')
# get lastfit attributes
par = {pn: getattr(lastfit, pn) for pn in lastfit.attr}
fig.suptitle('Fit to model ' + par.pop('func_name', '--'))
txt = ''
txt += f'$\chi^2$ = {par.pop("chi2", -1):.4G}\n'
txt += f'dof = {par.pop("dof", -1):.4G}\n'
txt += f'$cov_{{max}}$ = {np.max(par.pop("cov", 0)):.4G}\n'
_ = par.pop('func_code', None)
_ = par.pop('@name', None)
_ = par.pop('comment', None)
txtfree = ''
txtfix = ''
txtadd = ''
for attr in (p for p in par if p[-4:] != '_err'):
val = par.get(attr, None)
if not isinstance(val, numbers.Number):
txtadd += f'{attr} {val}\n'
continue
err = par.get(attr + '_err', None)
if err is None:
txtfix += f'{attr:<10} = {val:.3G}\n'
else:
txtfree += f'{attr:<10} = {val:.3G} ± {err:.3G}\n'
txtkwargs.pop('text', None)
txtkwargs.pop('s',
None) # in some places in matplotlib it still 's' instead of text, positional its at 3rd position
txtkwargs.pop('x', None)
txtkwargs.pop('y', None)
ax3pos = ax3.get_position()
fig.text(ax2.get_position().xmin * 1.15,
(ax3pos.ymax + ax3pos.ymin) / 2,
txt + txtfree + txtfix + txtadd,
verticalalignment='center', # 'top'
transform=fig.transFigure, **txtkwargs)
show(block=False)
return fig
[docs]def plot2Dimage(data, shape=None, yaxis_label='Z', xaxis_label='X', method='nearest', colorMap='jet', scale='norm',
linthresh=1, linscale=1, badcolor=None, transpose=None, figsize=[6, 6], origin='upper',
txtkwargs={'fontsize': 'small'}):
"""
Show a 2D image of a dataarray with XZW values like from oriented cloudscattering.
Parameters
----------
data : dataArray
dataArray optional with fit values in lastfit.
shape : [int,int]
Optional shape of the data if these are from an image.
If not given the data are interpolated (regrid)
yaxis_label : string
xaxis_label : string
method : float,'linear', 'nearest', 'cubic'
Filling value for new points as float or order of interpolation
between existing points.
See `griddata <https://docs.scipy.org/doc/scipy/reference/generated/scipy.interpolate.griddata.html>`_
colorMap : string
Get a colormap instance from name.
Standard mpl colormap name (see showColors).
scale : 'log', 'symlog', default = 'norm'
Scale for intensities.
- 'norm' Linear scale.
- 'log' Logarithmic scale
- 'symlog' Symmetrical logarithmic scale is logarithmic in both the positive
and negative directions from the origin. This works also for only positive data.
Use linthresh, linscale to adjust.
linthresh : float, default = 1
Only used for scale 'sym'.
The range within which the plot is linear (-linthresh to linthresh).
linscale : float, default = 1
Only used for scale 'sym'.
Its value is the number of decades to use for each half of the linear range.
E.g. 10 uses 1 decade.
badcolor : float, color
Set the color for bad values (like masked pixel) values in an image.
Default is bad values be transparent.
Color can be matplotlib color as 'k','b' or
float value in interval [0,1] of the chosen colorMap.
0 sets to minimum value, 1 to maximum value.
transpose : bool
Transpose coordinates, e.g. for sasImages.
figsize : [float,float], default [6,6]
Figure Size in inch.
txtkwargs : kwargs
Keyword arguments passed to Text https://matplotlib.org/api/text_api.html#matplotlib.text.Text).
except x,y,text arguments.
Returns
-------
figure
Examples
--------
::
import jscatter as js
import numpy as np
R=8 # maximum
N=200 # number of points
ds=15;
qxy=np.mgrid[-R:R:N*1j, -R:R:N*1j].reshape(2,-1).T
# add z=0 component
qxyz=np.c_[qxy,np.zeros(qxy.shape[0])].T # as position vectors
# create fcc lattice which includes reciprocal lattice vectors and methods to get peak positions
fcclattice= js.lattice.fccLattice(5, 5)
# Orient 111 direction perpendicular to qxy plane
fcclattice.rotatehkl2Vector([1,1,1],[0,0,1])
# rotation by 15 degrees to be aligned to xy plane
fcclattice.rotateAroundhkl([1,1,1],np.deg2rad(15))
ffs=js.sf.orientedLatticeStructureFactor(qxyz,fcclattice, rotation=[1,1,1,np.deg2rad(10)],
domainsize=ds,rmsd=0.1,hklmax=5,nGauss=23)
js.mpl.plot2Dimage(ffs)
.. image:: ../../examples/images/2dfccplot.png
:align: center
:width: 70 %
:alt: 2dfccplot
"""
# use copy so that we do not mutate the global colormap instance; stupid matplotlib programmers
cmap = copy.copy(matplotlib.cm.get_cmap(colorMap))
# if hasattr(data, 'W') and shape is None:
# print('This function does not yet handle 3D data!')
# return
if badcolor is not None:
# set bad color
if isinstance(badcolor, numbers.Number):
cmap.set_bad(color=cmap(badcolor))
else:
cmap.set_bad(color=badcolor)
# determine the scaling (norm)
# determine vmin,vmax later
if scale[:3] == 'log':
norm1 = colors.LogNorm(clip=True)
elif scale[:3] == 'sym':
norm1 = colors.SymLogNorm(clip=True, linthresh=linthresh, linscale=linscale)
else: # default: scale == 'normalize':
norm1 = colors.Normalize(clip=True)
norm2 = colors.Normalize(clip=True)
if shape is not None:
X, Z, Y = regrid(data.X, data.Z, data.Y, shape)
extend = [np.min(Z), np.max(Z), np.min(X), np.max(X)]
else:
X, Z, Y = regrid(data.X, data.Z, data.Y)
extend = [np.min(Z), np.max(Z), np.min(X), np.max(X)]
if transpose:
Y = Y.T
extend = [extend[2], extend[3], extend[0], extend[1]]
fig = figure(figsize=figsize)
ax1 = fig.add_subplot(1, 1, 1)
# add images and and some text
norm1.autoscale(Y)
im1 = ax1.imshow(Y, cmap=cmap, norm=norm1, extent=extend, origin=origin)
fig.colorbar(im1, ax=ax1, pad=0.01, shrink=0.85)
ax1.set_ylabel(yaxis_label)
ax1.set_xlabel(xaxis_label)
# get lastfit attributes
show(block=False)
return fig
[docs]def test(keepopen=True):
"""
A small test for mpl module making a plot.
Examples
--------
::
import jscatter as js
import numpy as np
from matplotlib import pyplot
# use this
#fig=pyplot.figure(FigureClass=js.mpl.Figure)
# or
fig=js.mplot()
fig.Multi(2,1)
fig[0].SetView(0.1,0.25,0.8,0.9)
fig[1].SetView(0.1,0.09,0.8,0.23)
q=js.loglist(0.01,5,100)
aa=js.dL()
for pp in range(5):
aa.append(js.dA(np.c_[q,-pp*np.sin(q),0.2*np.cos(5*q)].T))
aa[-1].qq=pp
bb=js.dA(np.c_[q,q**2].T)
bb.qq=123
for pp in range(5):
fig[0].Plot(aa[pp].X,-1*aa[pp].Y,legend='some stufff',sy=[1,(pp+1)/10.],li=0)
fig[0].Plot(aa, legend='qq = $qq', sy=[-1, 0.4, -1, ''], li=0, markeredgewidth=1)
for pp in range(5):
fig[1].Plot(aa[-1].X/5+pp,pp*aa[-1].Y,legend='q=%.1f' %pp,sy=0,li=-1,markeredgewidth =1)
fig[1].Plot(bb,legend='sqr=$qq ',sy=2,li=2)
fig[0].Title('test')
fig[0].Legend(x=1.3,y=1)
fig[1].Legend(x=1.3,y=1)
fig[0].Yaxis(label='y-axis')
fig[1].Yaxis(label='something else')
fig[0].tick_params(labelbottom=False)
fig[1].Xaxis(label='x-axis')
"""
import jscatter as js
import numpy as np
# use this
# fig=js.mpl.figure(FigureClass=js.mpl.Figure)
# or
fig1 = js.mplot()
fig1.Multi(2, 1)
fig1[0].SetView(0.1, 0.25, 0.8, 0.9)
fig1[1].SetView(0.1, 0.09, 0.8, 0.23)
q = js.loglist(0.01, 5, 100)
aa = js.dL()
for pp in range(5):
aa.append(js.dA(np.c_[q, -pp * np.sin(q), 0.2 * np.cos(5 * q)].T))
aa[-1].qq = pp
bb = js.dA(np.c_[q, q ** 2].T)
bb.qq = 123
for pp in range(5):
fig1[0].Plot(aa[pp].X, -1 * aa[pp].Y, legend='some stufff', sy=[1, (pp + 1) / 10.], li=0)
fig1[0].Plot(aa, legend='qq = $qq', sy=[-1, 0.4, -1, ''], li=0, markeredgewidth=1)
for pp in range(5):
fig1[1].Plot(aa[-1].X / 5 + pp, pp * aa[-1].Y, legend='q=%.1f' % pp, sy=0, li=-1, markeredgewidth=1)
fig1[1].Plot(bb, legend='sqr=$qq ', sy=2, li=2)
fig1[0].Title('test')
fig1[0].Legend(x=1.3, y=1)
fig1[1].Legend(x=1.3, y=1)
fig1[0].Yaxis(label='y-axis')
fig1[1].Yaxis(label='something else')
fig1[0].tick_params(labelbottom=False)
fig1[1].Xaxis(label='x-axis')
fig1.savefig('mpltest1.png')
import jscatter as js
import numpy as np
calibration = js.sas.sasImage(js.examples.datapath + '/calibration.tiff')
fig2 = js.mpl.contourImage(np.ma.log(calibration))
fig2.savefig('mpltest2.png')
x, z = np.mgrid[-5:5:0.25, -5:5:0.25]
xyz = js.dA(np.c_[x.flatten(), z.flatten(), 0.3 * np.sin(x * z / np.pi).flatten() + 0.01 * np.random.randn(
len(x.flatten())), 0.01 * np.ones_like(x).flatten()].T)
xyz.setColumnIndex(ix=0, iy=2, iz=1)
fig3 = js.mpl.contourImage(xyz)
fig3.savefig('mpltest3.png')
# random distributed points
x, z = js.formel.randomPointsInCube(1500, 0, 2).T * 10 - 5
xyz = js.dA(np.c_[x.flatten(), z.flatten(), 0.3 * np.sin(x * z / np.pi).flatten() + 0.01 * np.random.randn(
len(x.flatten()))].T)
xyz.setColumnIndex(ix=0, iy=2, iz=1)
fig4 = js.mpl.contourImage(xyz)
fig4.savefig('mpltest4.png')
if keepopen:
return fig4
else:
js.mpl.pyplot.close('all')