def make_1d_moments(outdir = '', sim_type = 'tristan-mp', n='1', prtl_type='',
xval='', yval = '', weights = '', boolstr = '', xbins ='200', xvalmin = '',
xvalmax = '', xtra_stride = '1', xscale = 'linear',
selPolyXval = '', selPolyYval = '', selPolyXarr = '', selPolyYarr= ''):
'''We calculate a 1D average of yval as a function of xval at outdir,
and then return a list of dictionaries
where each dictinary is of the form
{'num': the number count,
'x0': left edge of the bin,
'x1': right edge of the bin}.'''
### first we open up a tristan sim
if sim_type =='tristan-mp':
cur_sim = TristanSim(outdir, n = int(n), xtra_stride = int(xtra_stride))
# first we evaluate the boolean string to see what values we should discard:
bool_arr = parse_boolstr(boolstr, cur_sim, prtl_type)
# Now we go through an fill out some of the unfilled data needed to make
# take an average
xarr = getattr(getattr(cur_sim, prtl_type), xval)
xarr = xarr if bool_arr is None else xarr[bool_arr]
yarr = getattr(getattr(cur_sim, prtl_type), yval)
yarr = yarr if bool_arr is None else yarr[bool_arr]
warr = np.array([])
if len(weights) != 0:
warr = getattr(getattr(cur_sim, prtl_type), weights)
warr = warr if bool_arr is None else warr[bool_arr]
# NOW WE APPLY THE POLYGON:
if (len(selPolyXval) != 0):
inside = np.zeros(len(xarr), dtype='bool')
datX = getattr(getattr(cur_sim, prtl_type), selPolyXval)
datX = datX if bool_arr is None else datX[bool_arr]
datY = getattr(getattr(cur_sim, prtl_type), selPolyYval)
datY = datY if bool_arr is None else datY[bool_arr]
polyX = np.fromstring(selPolyXarr, sep=',')
polyY = np.fromstring(selPolyYarr, sep=',')
bbox = np.array([polyX.min(), polyX.max(), polyY.min(), polyY.max()])
point_in_polygon(datX, datY, polyX, polyY, bbox, inside)
xarr = xarr[inside]
yarr = yarr[inside]
if (len(warr) != 0):
warr = warr[inside]
xvalmin = xarr.min() if len(xvalmin)==0 else float(xvalmin)
xvalmax = xarr.max() if len(xvalmax)==0 else float(xvalmax)
if len(warr)==0:
if xscale =='log' and xvalmin > 0:
hist = CalcMoments(np.log10(xarr), yarr, np.log10(xvalmin), np.log10(xvalmax), int(float(xbins)))
else:
hist = CalcMoments(xarr, yarr, xvalmin, xvalmax, int(float(xbins)))
else:
#calculate unweighed histogram
if xscale =='log' and xvalmin > 0:
hist = CalcWeightedMoment(np.log10(xarr), yarr, warr, np.log10(xvalmin), np.log10(xvalmax), int(float(xbins)))
else:
hist = CalcWeightedMoment(xarr, yarr, warr, xvalmin, xvalmax, int(float(xbins)))
####
#
# Now we have the histogram, we need to turn it into a JSON compatible with
# D3 hist function. D3 hist return a sorted array that keeps all of the particle
# data. That is not possible in our case due to memory constraints. Instead we
# simply return a list
#
###
if xscale =='log' and xvalmin >0:
bin_width = (np.log10(xvalmax)-np.log10(xvalmin))/int(xbins)
bins = np.logspace(np.log10(xvalmin), np.log10(xvalmax), num = int(xbins)+1)
else:
bin_width = (xvalmax-xvalmin)/int(xbins)
bins = np.linspace(xvalmin, xvalmax, num = int(xbins)+1)
x_arr, y_arr = stepify(bins, hist)
x_arr = x_arr[1:-1]; y_arr = y_arr[1:-1]
mom1D = [{'y': y_arr[i],
'x': x_arr[i]} for i in range(len(x_arr))]
return {
'lineData': mom1D,
'xscale': 'log' if xscale =='log' and xvalmin >0 else 'linear',
'xmin': mom1D[0]['x'],
'xmax': mom1D[-1]['x'],
'vmin': np.min(hist),
'vmax': np.max(hist)
}
def make_2d_mom_img(outdir = '', sim_type = 'tristan-mp', n='1', prtl_type='',
yval='', xval='', mval='', weights = '', boolstr = '', ybins = '200',
xbins ='200', yvalmin='', yvalmax='', xvalmin = '',
xvalmax = '', normhist = 'true',cmap='viridis', cnorm = 'log',
pow_zero = '0', pow_gamma='1.0', vmin = '', clip = 'true',
vmax = '', xmin='', xmax ='', ymin='', ymax='', interpolation = 'bicubic',
px ='400', py='400', aspect='auto', mask_zeros='true', xtra_stride = '1',
selPolyXval= '', selPolyYval='', selPolyXarr='', selPolyYarr=''):
# First we calculate the histogram, then we turn it into an image and return
# the image as a bytesIO
### first we open up a tristan sim
if sim_type =='tristan-mp':
cur_sim = TristanSim(outdir, n = int(n), xtra_stride = int(xtra_stride))
# first we evaluate the boolean string to see what values we should discard:
bool_arr = parse_boolstr(boolstr, cur_sim, prtl_type)
# Now we go through an fill out some of the unfilled data needed to make
# a histogram
xarr = getattr(getattr(cur_sim, prtl_type), xval)
xarr = xarr if bool_arr is None else xarr[bool_arr]
yarr = getattr(getattr(cur_sim, prtl_type), yval)
yarr = yarr if bool_arr is None else yarr[bool_arr]
marr = getattr(getattr(cur_sim, prtl_type), mval)
marr = marr if bool_arr is None else marr[bool_arr]
warr = np.array([])
if len(weights) != 0:
warr = getattr(getattr(cur_sim, prtl_type), weights)
warr = warr if bool_arr is None else warr[bool_arr]
# NOW WE APPLY THE POLYGON:
if (len(selPolyXval) != 0):
inside = np.zeros(len(xarr), dtype='bool')
datX = getattr(getattr(cur_sim, prtl_type), selPolyXval)
datX = datX if bool_arr is None else datX[bool_arr]
datY = getattr(getattr(cur_sim, prtl_type), selPolyYval)
datY = datY if bool_arr is None else datY[bool_arr]
polyX = np.fromstring(selPolyXarr, sep=',')
polyY = np.fromstring(selPolyYarr, sep=',')
bbox = np.array([polyX.min(), polyX.max(), polyY.min(), polyY.max()])
point_in_polygon(datX, datY, polyX, polyY, bbox, inside)
xarr = xarr[inside]
yarr = yarr[inside]
marr = marr[inside]
if (len(warr) != 0):
warr = warr[inside]
yvalmin = yarr.min() if len(yvalmin)==0 else float(yvalmin)
yvalmax = yarr.max() if len(yvalmax)==0 else float(yvalmax)
xvalmin = xarr.min() if len(xvalmin)==0 else float(xvalmin)
xvalmax = xarr.max() if len(xvalmax)==0 else float(xvalmax)
if len(warr)==0:
hist = Calc2DMoments(yarr, xarr, marr, yvalmin, yvalmax, int(float(ybins)), xvalmin, xvalmax, int(float(xbins)))
else:
#calculate unweighed histogram
hist = Calc2DWeightedMoments(yarr, xarr, marr, warr, yvalmin, yvalmax, int(float(ybins)), xvalmin, xvalmax, int(float(xbins)))
####
#
# Now we have the histogram, we need to turn it into an image
#
###
hist_img = myNumbaImage(int(py), int(px))
hist_img.setInterpolation(interpolation)
hist_img.setData(hist)
hist_img.setExtent([xvalmin,xvalmax,yvalmin, yvalmax])
hist_img.set_xlim(xmin = None if len(xmin)==0 else float(xmin),
xmax = None if len(xmax)==0 else float(xmax))
hist_img.set_ylim(ymin = None if len(ymin)==0 else float(ymin),
ymax = None if len(ymax)==0 else float(ymax))
if cnorm =='log':
hist_img.setNorm('log', clipped = True if clip =='true' else False)
if cnorm =='linear':
hist_img.setNorm('linear', clipped = True if clip =='true' else False)
if cnorm =='pow':
hist_img.setNorm('pow',zero = float(pow_zero), gamma = float(pow_gamma), clipped = True if clip =='true' else False)
hist_img.setCmap(cmap)
hist_img.set_clim(cmin = None if len(vmin) ==0 else float(vmin), cmax = None if len(vmax)==0 else float(vmax))
hist_img.set_aspect(0 if aspect=='auto' else 1)
#hist_img.set_aspect(1)# if aspect=='auto' else 1)
return hist_img.renderImageDict()
xmax=None if xmax == '' else float(xmax))
hist_img.set_ylim(ymin=None if ymin == '' else float(ymin),
ymax=None if ymax == '' else float(ymax))
if cnorm == 'log':
hist_img.setNorm('log', clipped=True if clip == True else False)
if cnorm == 'linear':
hist_img.setNorm('linear', clipped=True if clip == True else False)
if cnorm == 'pow':
hist_img.setNorm('pow',
zero=float(pow_zero),
gamma=float(pow_gamma),
clipped=True if clip == 'true' else False)
hist_img.setCmap(cmap)
hist_img.set_clim(cmin=None if vmin == '' else float(vmin),
cmax=None if vmax == '' else float(vmax))
hist_img.set_aspect(0 if aspect == 'auto' else 1)
#hist_img.set_aspect(1)# if aspect=='auto' else 1)
return hist_img.renderImageDict()
if __name__ == '__main__':
import numpy as np
from tristan_sim import TristanSim
mySim = TristanSim('../test_output')
print(
np.all(
np.where(parse_boolstr('[x.gt.20.0]AND[x.lt.30]', mySim, 'ions'))
[0] == np.where(
parse_boolstr('[x.gt.20]AND[x.lt.30]', mySim, 'ions'))[0]))
print(mySim.ions.quantities)