def constrainASCII(pathascii, filename, data):
(ncols, nrows, xllcorner, yllcorner, cellsize,
NODATA_value) = wk.readheader(pathascii, filename)
# take the minimum of ncols, nrows and ncols and nrows of data
nrowsd, ncolsd = np.shape(data)
ncolsnrows = np.min([ncols, nrows, ncolsd, nrowsd])
data = data[:ncolsnrows, :ncolsnrows]
print(ncols, nrows, ncolsd, nrowsd)
x = np.linspace(0, (ncolsnrows - 1) * cellsize, ncolsnrows)
y = np.linspace(0, (ncolsnrows - 1) * cellsize, ncolsnrows)
xe = np.linspace(cellsize / 2.0,
(cellsize / 2.0) + ((ncolsnrows - 1) * cellsize),
ncolsnrows)
ye = np.linspace(cellsize / 2.0,
(cellsize / 2.0) + ((ncolsnrows - 1) * cellsize),
ncolsnrows)
# create my own matrices for the plotting with pcolor or pcolormesh
xc = np.zeros_like(data)
yc = np.zeros_like(data)
xce = np.zeros_like(data)
yce = np.zeros_like(data)
for i in range(ncolsnrows):
xc[:, i] = x
xce[:, i] = xe
for i in range(ncolsnrows):
yc[i, :] = y
yce[i, :] = ye
# middle of the dtm (this has to change!)
ncentery = int(nrows / 2) # changed
ncenterx = int(ncols / 2) # changed
return (xc, yc, data, ncenterx, ncentery)
def run1(path, filenameXY, filenamecrater, pathplot, pathdata):
'''
filenameXY = 'data.txt'
filenamecrater = 'crater_id.txt'
'''
# the ASCII file (converted from the raster in ArcGIS) is loaded
os.chdir(path)
#filenames = glob.glob("CPcrater*.asc") #need to fix that if coldspots are included?
#filenames.sort(key=wk.tokenize)
dataXYD = np.loadtxt(filenameXY,delimiter=";",comments="#")
crater_id = np.genfromtxt(filenamecrater,skip_header=1,dtype=str)
xcrater = dataXYD[:,0]
ycrater = dataXYD[:,1]
diam0 = dataXYD[:,2]
diam0 = diam0 * 1000.
r0 = diam0 /2.
#it takes way too much time this way
#xarcout = np.array([])
#yarcout = np.array([])
#if it is the last step
for indf, filename in enumerate(crater_id):
print (indf)
# should be good this way
data = np.loadtxt(path + filename + ".asc", skiprows=6)
data2 = np.rot90(data)
data3 = np.rot90(data2)
data4 = np.rot90(data3)
del data
del data2
del data3
data = copy.deepcopy(data4) #multiplying factor of the elevation (not for kaguya, 0.5 for Kaguya + LOLA)
del data4
#correspond to orgin in the lower-left corner and data[x,y]
# with x = ncols, y = nrows !!! important
# the number of columns, rows and extent of the raster is read
(ncols, nrows, xllcorner, yllcorner, cellsize, NODATA_value) = wk.readheader(path,filename + ".asc")
# take the minimum of ncols, nrows and ncols and nrows of data
nrowsd, ncolsd = np.shape(data)
ncolsnrows = np.min([ncols, nrows, ncolsd, nrowsd])
data = data[:ncolsnrows, :ncolsnrows]
x = np.linspace(0,(ncolsnrows-1)*cellsize,ncolsnrows)
y = np.linspace(0,(ncolsnrows-1)*cellsize,ncolsnrows)
xe = np.linspace(cellsize/2.0, (cellsize/2.0) + ((ncolsnrows-1)*cellsize), ncolsnrows)
ye = np.linspace(cellsize/2.0, (cellsize/2.0) + ((ncolsnrows-1)*cellsize), ncolsnrows)
# create my own matrices for the plotting with pcolor or pcolormesh
xc = np.zeros_like(data)
yc = np.zeros_like(data)
xce = np.zeros_like(data)
yce = np.zeros_like(data)
for i in range(ncolsnrows):
xc[:,i] = x
xce[:,i] = xe
for i in range(ncolsnrows):
yc[i,:] = y
yce[i,:] = ye
# middle of the dtm (this has to change!)
ncentery = int(nrows/2) # changed
ncenterx = int(ncols/2) # changed
#other places where ncenterx and ncentery is required
x1, y1 = wk.xy_circle(1.0*r0[indf], xe[ncenterx], ye[ncentery]) # xe and ye are equals
# First detrending (a plane is fitted through the elevations taken at circles
# 2.0 and 3.0R from the center of the crater )
#stduse = True #use standard deviation removal
#ndata = wk.detrending(xc, yc, r0[indf], cellsize, ncenterx, ncentery, data, stduse) # I think it's okay to use xc, yc here
# txt name
name_crater_txt = filename + 'XY.txt'
#and if XY is not found in a certain path
if os.path.isfile(pathdata + name_crater_txt):
pass
else:
# the Maximum elevation are selected a first time
#first_run = True
mingrade = 0.05
minclust = 0.05
slen = 0.1
#(col_coord_ME, row_coord_ME, col_cells_ME, row_cells_ME, elev_ME, prof_ME) = (
# wk.rim(xc, yc, ncenterx, ncentery, ndata, r0[indf],
# cellsize, slen, minclust, mingrade, first_run))
# we get rid of the nan-values for the detrending
#ixnan = np.where(np.isnan(col_coord_ME) == False)
# Maximum elevations without NaNs
#coldetrend = col_coord_ME[ixnan]
#rowdetrend = row_coord_ME[ixnan]
#elev4detrend = elev_ME[ixnan]
# the second detrending through the selected maximum elevation points are done
#stduse = False
#Z2 = wk.linear3Ddetrending(coldetrend,rowdetrend,elev4detrend, xc, yc, stduse)
# the detrended plane is substracted to the DEM
#ndata2 = ndata - Z2
# the first detrending has ben run so we change now the first_run flag to false
first_run = False
# second run of the function rim
(col_coord_ME, row_coord_ME, col_cells_ME, row_cells_ME, elev_ME, prof_ME,
col_coord_LE, row_coord_LE, col_cells_LE, row_cells_LE, elev_LE, prof_LE,
col_coord_BS, row_coord_BS, col_cells_BS, row_cells_BS, elev_BS, prof_BS) = (
wk.rim(xc, yc, ncenterx, ncentery, data, r0[indf],
cellsize, slen, minclust, mingrade, first_run))
# takes only nonnan values from *BS data and merge it to local elevation data
ixfinite = np.where(np.isfinite(col_coord_BS) == True)
col_coord_BS = col_coord_BS[ixfinite]
row_coord_BS = row_coord_BS[ixfinite]
col_cells_BS = col_cells_BS[ixfinite]
row_cells_BS = row_cells_BS[ixfinite]
elev_BS = elev_BS[ixfinite]
prof_BS = prof_BS[ixfinite]
use_break_in_slope = False
if use_break_in_slope:
# concatenate local maxima and slope change!
colint = np.concatenate((col_coord_LE, col_coord_BS))
rowint = np.concatenate((row_coord_LE, row_coord_BS))
colmap = np.concatenate((col_cells_LE, col_cells_BS))
rowmap = np.concatenate((row_cells_LE, row_cells_BS))
elevint = np.concatenate((elev_LE, elev_BS))
profint = np.concatenate((prof_LE, prof_BS))
else:
# or using only local maxima
colint = col_coord_LE
rowint = row_coord_LE
colmap = col_cells_LE
rowmap = row_cells_LE
elevint = elev_LE
profint = prof_LE
# Maximum allowed radial discontinuity Drad (I should convert these values in cells)
Drad = 0.1 * r0[indf]
# Distance of interest (searching distance)
Dint = 0.05 * r0[indf]
# Maximum angular discontinuity (avoid unnecessary large gap angle in the
# data)
angle = 2.0 #(in degrees)
stangle = [0,45,90,135,180,225,270,315]
contloop = True
siftRedundant = True
kpstitch = False
OptRims, Omegas, gap, maxradf = wk.rim_composite(col_coord_ME, row_coord_ME, col_cells_ME,
row_cells_ME, elev_ME, prof_ME,
colint, rowint, colmap, rowmap, xc, yc, elevint,
profint, ncenterx, ncentery,
angle, stangle, Drad, Dint,
contloop, siftRedundant, kpstitch)
# we want to export the data as X, Y, Z
# find the best fit
for i in range(np.shape(OptRims)[0]):
a = np.mean(maxradf[i]) + (0.5*Omegas[i])
if i == 0:
b = a
c = i
else:
if a < b:
b = a
c = i
# I should get rid of the 0,0 counts and should not append it takes way too much time
# temporary name
name_crater_png = filename + 'XY.png'
#xarcout = np.append(xarcout, np.array(OptRims[c][0,:]) + xllcorner)
#yarcout = np.append(yarcout, np.array(OptRims[c][1,:]) + yllcorner)
plt.ioff()
plt.figure()
plt.pcolormesh(xc, yc, data)
plt.title(filename,fontsize=16)
plt.colorbar()
xarcgis = np.array(OptRims[c][0,:])
xarcgis = xarcgis[np.nonzero(xarcgis)]
yarcgis = np.array(OptRims[c][1,:])
yarcgis = yarcgis[np.nonzero(yarcgis)]
zarcgis = np.array(OptRims[c][2,:])
zarcgis = zarcgis[np.nonzero(yarcgis)]
profgis = np.array(OptRims[c][3,:])
profgis = profgis[np.nonzero(yarcgis)]
flaggis = np.array(OptRims[c][4,:])
flaggis = flaggis[np.nonzero(yarcgis)]
#xarcgis = np.array(OptRims[c][0,:]) + xllcorner
#xarcgis = xarcgis[np.nonzero(xarcgis - xllcorner)]
#yarcgis = np.array(OptRims[c][1,:]) + yllcorner
#yarcgis = yarcgis[np.nonzero(yarcgis - yllcorner)]
#zarcgis = np.array(OptRims[c][2,:])
#zarcgis = zarcgis[np.nonzero(yarcgis - yllcorner)]
#profgis = np.array(OptRims[c][3,:])
#profgis = profgis[np.nonzero(yarcgis - yllcorner)]
#flaggis = np.array(OptRims[c][4,:])
#flaggis = flaggis[np.nonzero(yarcgis - yllcorner)]
#print np.shape(OptRims)
plt.plot(xarcgis,yarcgis,"ko")
plt.plot(x1,y1,"b")
#fit a circle
xnewcenter, ynewcenter, rnew, residu = wk.leastsq_circle(xarcgis,yarcgis)
#other places where ncenterx and ncentery is required (new calculation)
x1n, y1n = wk.xy_circle(1.0*rnew, xnewcenter, ynewcenter)
plt.plot(x1n,y1n,"r")
# save plot
plt.savefig(pathplot + name_crater_png,dpi=300)
plt.close()
#export the data to a .csv file with ";"
#xarcgis = np.array(OptRims[c][0,:]) + xllcorner
#yarcgis = np.array(OptRims[c][1,:]) + yllcorner
#zarcgis = np.array(OptRims[c][2,:])
header_txt = 'X;Y;Z;prof;flag'
arcout = np.column_stack((xarcgis, yarcgis, zarcgis, profgis, flaggis))
np.savetxt(pathdata + name_crater_txt, arcout, delimiter = ";", header=header_txt,fmt='%10.5f', comments='')
def run2(path, filenameXY, filenamecrater, pathdata):
'''
filenameXY = 'data.txt'
filenamecrater = 'crater_id.txt'
'''
os.chdir(path)
#filenames = glob.glob("CPcrater*.asc")
#filenames.sort(key=wk.tokenize)
crater_id = np.genfromtxt(filenamecrater,skip_header=1,dtype=str)
#dataXYD = np.loadtxt(filenameXY,delimiter=";",comments="#")
#xcrater = dataXYD[:,0]
#ycrater = dataXYD[:,1]
#diam0 = dataXYD[:,2]
#diam0 = diam0 * 1000.
#r0 = diam0 /2.
unc_cse = np.ones(len(crater_id))
med_cse = np.ones(len(crater_id))
cse_25 = np.ones(len(crater_id))
cse_75 = np.ones(len(crater_id))
cse_min = np.ones(len(crater_id))
cse_max = np.ones(len(crater_id))
unc_mcw = np.ones(len(crater_id))
med_mcw = np.ones(len(crater_id))
mcw_25 = np.ones(len(crater_id))
mcw_75 = np.ones(len(crater_id))
mcw_min = np.ones(len(crater_id))
mcw_max = np.ones(len(crater_id))
unc_ucw = np.ones(len(crater_id))
med_ucw = np.ones(len(crater_id))
ucw_25 = np.ones(len(crater_id))
ucw_75 = np.ones(len(crater_id))
ucw_min = np.ones(len(crater_id))
ucw_max = np.ones(len(crater_id))
unc_h = np.ones(len(crater_id))
med_h = np.ones(len(crater_id))
h_25 = np.ones(len(crater_id))
h_75 = np.ones(len(crater_id))
h_min = np.ones(len(crater_id))
h_max = np.ones(len(crater_id))
med_depth = np.ones(len(crater_id))
depthf = np.ones(len(crater_id))
depth_25 = np.ones(len(crater_id))
depth_75 = np.ones(len(crater_id))
depth_min = np.ones(len(crater_id))
depth_max = np.ones(len(crater_id))
diamf = np.ones(len(crater_id))
med_diam = np.ones(len(crater_id))
diam_25 = np.ones(len(crater_id))
diam_75 = np.ones(len(crater_id))
diam_min = np.ones(len(crater_id))
diam_max = np.ones(len(crater_id))
unc_crdl = np.ones(len(crater_id))
med_crdl = np.ones(len(crater_id))
crdl_25 = np.ones(len(crater_id))
crdl_75 = np.ones(len(crater_id))
crdl_min = np.ones(len(crater_id))
crdl_max = np.ones(len(crater_id))
unc_frdl = np.ones(len(crater_id))
med_frdl = np.ones(len(crater_id))
frdl_25 = np.ones(len(crater_id))
frdl_75 = np.ones(len(crater_id))
frdl_min = np.ones(len(crater_id))
frdl_max = np.ones(len(crater_id))
unc_rupcw = np.ones(len(crater_id))
med_rupcw = np.ones(len(crater_id))
rupcw_25 = np.ones(len(crater_id))
rupcw_75 = np.ones(len(crater_id))
rupcw_min = np.ones(len(crater_id))
rupcw_max = np.ones(len(crater_id))
unc_rufrc = np.ones(len(crater_id))
med_rufrc = np.ones(len(crater_id))
rufrc_25 = np.ones(len(crater_id))
rufrc_75 = np.ones(len(crater_id))
rufrc_min = np.ones(len(crater_id))
rufrc_max = np.ones(len(crater_id))
unc_lrs = np.ones(len(crater_id))
med_lrs = np.ones(len(crater_id))
lrs_25 = np.ones(len(crater_id))
lrs_75 = np.ones(len(crater_id))
lrs_min = np.ones(len(crater_id))
lrs_max = np.ones(len(crater_id))
unc_urs = np.ones(len(crater_id))
med_urs = np.ones(len(crater_id))
urs_25 = np.ones(len(crater_id))
urs_75 = np.ones(len(crater_id))
urs_min = np.ones(len(crater_id))
urs_max = np.ones(len(crater_id))
unc_fsa = np.ones(len(crater_id))
med_fsa = np.ones(len(crater_id))
fsa_25 = np.ones(len(crater_id))
fsa_75 = np.ones(len(crater_id))
fsa_min = np.ones(len(crater_id))
fsa_max = np.ones(len(crater_id))
for indf, filename in enumerate(crater_id):
print (indf)
# load data
data = np.loadtxt(path + filename + ".asc", skiprows=6)
data2 = np.rot90(data)
data3 = np.rot90(data2)
data4 = np.rot90(data3)
del data
del data2
del data3
data = copy.deepcopy(data4) # multiplying by scaling factor (*0.50 for Kaguya+LOLA)
del data4
# txt name
name_crater_txt = filename + 'XY.txt'
# the number of columns, rows and extent of the raster is read
(ncols, nrows, xllcorner, yllcorner, cellsize, NODATA_value) = wk.readheader(path, filename + ".asc")
# take the minimum of ncols, nrows and ncols and nrows of data
nrowsd, ncolsd = np.shape(data)
ncolsnrows = np.min([ncols, nrows, ncolsd, nrowsd])
data = data[:ncolsnrows, :ncolsnrows]
x = np.linspace(0,(ncolsnrows-1)*cellsize,ncolsnrows)
y = np.linspace(0,(ncolsnrows-1)*cellsize,ncolsnrows)
xe = np.linspace(cellsize/2.0, (cellsize/2.0) + ((ncolsnrows-1)*cellsize), ncolsnrows)
ye = np.linspace(cellsize/2.0, (cellsize/2.0) + ((ncolsnrows-1)*cellsize), ncolsnrows)
# create my own matrices for the plotting with pcolor or pcolormesh
xc = np.zeros_like(data)
yc = np.zeros_like(data)
xce = np.zeros_like(data)
yce = np.zeros_like(data)
for i in range(ncolsnrows):
xc[:,i] = x
xce[:,i] = xe
for i in range(ncolsnrows):
yc[i,:] = y
yce[i,:] = ye
#load the data
datagis = loadgis(pathdata, name_crater_txt)
xarcgis = datagis[:,0]
yarcgis = datagis[:,1]
zarcgis = datagis[:,2]
profgis = datagis[:,3]
flaggis = datagis[:,4]
# new get rid of nan values
xarcgis = xarcgis[~np.isnan(xarcgis)]
yarcgis = yarcgis[~np.isnan(yarcgis)]
zarcgis = zarcgis[~np.isnan(zarcgis)]
profgis = profgis[~np.isnan(profgis)]
flaggis = flaggis[~np.isnan(flaggis)]
#get rid of outliers
#Q1 = np.percentile(zarcgis,25.)
#Q3 = np.percentile(zarcgis,75.)
#IQR = Q3 - Q1
#outliers = (zarcgis < (Q1 - 1.5 * IQR)) | (zarcgis > (Q3 + 1.5 * IQR))
#not_outliers = np.where(outliers == False)
x_not_outliers = xarcgis
y_not_outliers = yarcgis
z_not_outliers = zarcgis
prof_not_outliers = profgis
# old center
# get the cell the closest to the center of the crater
#__, ncenterx_old = wk.find_nearest(xllcorner+xe, xcrater[indf]) #correspond to columns and x
# because yllcorner is from the top left now
#__, ncentery_old = wk.find_nearest(yllcorner+ye,ycrater[indf]) #correspond to rows and y
#ncenterx_old = nrows/2
#ncentery_old = nrows/2
#fit a circle
xnewcenter, ynewcenter, rnew, residu = wk.leastsq_circle(x_not_outliers,y_not_outliers)
# get the cell the closest to the center of the crater (new calculation)
# (not good, use xnewcenter and ynewcenter)
__, ncenterx = wk.find_nearest(xe,xnewcenter)
__, ncentery = wk.find_nearest(ye,ynewcenter)
# new
#ncenterx = np.int(np.round(xnewcenter,decimals=0))
#ncentery = np.int(np.round(ynewcenter,decimals=0))
#other places where ncenterx and ncentery is required (new calculation)
#x1, y1 = wk.xy_circle(1.0*rnew, xe[ncentery], ye[ncenterx])
# First detrending (a plane is fitted through the elevations taken at circles
# with new center of the circle
try:
#stduse = True
# here change the center of the craters!
#ndata = wk.detrending(xc, yc, rnew, cellsize, ncenterx, ncentery, data, stduse) #detrending
#ndata = wk.detrending(xc, yc, rnew, cellsize, ncenterx_old, ncentery_old, data, stduse) #detrending
# with old data otherwise sometimes the dem area is not large enough
# the second detrending through the selected maximum elevation points are done
#stduse = False
#Z2 = wk.linear3Ddetrending(x_not_outliers,y_not_outliers,z_not_outliers, xc, yc, stduse)
#ndata2 = wk.detrending_rim(xc, yc, rnew, cellsize, ncenterx, ncentery, ndata, stduse)
# the detrended plane is substracted to the DEM (with the newly detected)
#ndata2 = ndata - Z2
# I should actually rerun the rim detection things one more time
# as the centre of the crater has been moved (not the same profiles anymore)
#need to put some indices there (maybe, they will have different sizes)
(R_upcw, R_ufrc, cse, slope_mcw, slope_ucw, slope_fsa, slope_lrs, slope_urs, crdl, frdl,
h, hr, depth, diam, nnn, prf, crossSections, YSections, XSections) = (wk.calculation(xc, yc, x_not_outliers,
y_not_outliers, z_not_outliers, prof_not_outliers,
rnew, data, cellsize, ncenterx, ncentery, xllcorner, yllcorner))
# crossSections has been added
# the problem is that there are several values for profile = 0? for some reasons?
# I don't understand that, I guess it must be profiles where....
# np.where(prf != 0) to get rid of it but it would be nice if the problem would be detected
#just need to make the calculation, we have the final circle
# need to find the closest values that does not overlap
unc_cse[indf] = np.nanstd(cse)/np.sqrt(nnn)
med_cse[indf] = np.nanmedian(cse)
cse_25[indf] = np.nanpercentile(cse,25)
cse_75[indf] = np.nanpercentile(cse,75)
cse_min[indf] = np.nanmin(cse)
cse_max[indf] = np.nanmax(cse)
unc_mcw[indf] = np.nanstd(slope_mcw)/np.sqrt(nnn)
med_mcw[indf] = np.nanmedian(slope_mcw)
mcw_25[indf] = np.nanpercentile(slope_mcw,25)
mcw_75[indf] = np.nanpercentile(slope_mcw,75)
mcw_min[indf] = np.nanmin(slope_mcw)
mcw_max[indf] = np.nanmax(slope_mcw)
unc_ucw[indf] = np.nanstd(slope_ucw)/np.sqrt(nnn)
med_ucw[indf] = np.nanmedian(slope_ucw)
ucw_25[indf] = np.nanpercentile(slope_ucw,25)
ucw_75[indf] = np.nanpercentile(slope_ucw,75)
ucw_min[indf] = np.nanmin(slope_ucw)
ucw_max[indf] = np.nanmax(slope_ucw)
unc_h[indf] = np.nanstd(h)/np.sqrt(nnn)
med_h[indf] = np.nanmedian(h)
h_25[indf] = np.nanpercentile(h,25)
h_75[indf] = np.nanpercentile(h,75)
h_min[indf] = np.nanmin(h)
h_max[indf] = np.nanmax(h)
unc_hr[indf] = np.nanstd(hr)/np.sqrt(nnn)
med_hr[indf] = np.nanmedian(hr)
hr_25[indf] = np.nanpercentile(hr,25)
hr_75[indf] = np.nanpercentile(hr,75)
hr_min[indf] = np.nanmin(hr)
hr_max[indf] = np.nanmax(hr)
depthf[indf] = np.nanstd(depth)/np.sqrt(nnn)
med_depth[indf] = np.nanmedian(depth)
depth_25[indf] = np.nanpercentile(depth,25)
depth_75[indf] = np.nanpercentile(depth,75)
depth_min[indf] = np.nanmin(depth)
depth_max[indf] = np.nanmax(depth)
diamf[indf] = np.nanstd(diam)/np.sqrt(nnn)
med_diam[indf] = np.nanmedian(diam)
diam_25[indf] = np.nanpercentile(diam,25)
diam_75[indf] = np.nanpercentile(diam,75)
diam_min[indf] = np.nanmin(diam)
diam_max[indf] = np.nanmax(diam)
##
unc_rupcw[indf] = np.nanstd(R_upcw)/np.sqrt(nnn)
med_rupcw[indf] = np.nanmedian(R_upcw)
rupcw_25[indf] = np.nanpercentile(R_upcw,25)
rupcw_75[indf] = np.nanpercentile(R_upcw,75)
rupcw_min[indf] = np.nanmin(R_upcw)
rupcw_max[indf] = np.nanmax(R_upcw)
unc_rufrc[indf] = np.nanstd(R_ufrc)/np.sqrt(nnn)
med_rufrc[indf] = np.nanmedian(R_ufrc)
rufrc_25[indf] = np.nanpercentile(R_ufrc,25)
rufrc_75[indf] = np.nanpercentile(R_ufrc,75)
rufrc_min[indf] = np.nanmin(R_ufrc)
rufrc_max[indf] = np.nanmax(R_ufrc)
unc_lrs[indf] = np.nanstd(slope_lrs)/np.sqrt(nnn)
med_lrs[indf] = np.nanmedian(slope_lrs)
lrs_25[indf] = np.nanpercentile(slope_lrs,25)
lrs_75[indf] = np.nanpercentile(slope_lrs,75)
lrs_min[indf] = np.nanmin(slope_lrs)
lrs_max[indf] = np.nanmax(slope_lrs)
unc_urs[indf] = np.nanstd(slope_urs)/np.sqrt(nnn)
med_urs[indf] = np.nanmedian(slope_urs)
urs_25[indf] = np.nanpercentile(slope_urs,25)
urs_75[indf] = np.nanpercentile(slope_urs,75)
urs_min[indf] = np.nanmin(slope_urs)
urs_max[indf] = np.nanmax(slope_urs)
unc_fsa[indf] = np.nanstd(slope_fsa)/np.sqrt(nnn)
med_fsa[indf] = np.nanmedian(slope_fsa)
fsa_25[indf] = np.nanpercentile(slope_fsa,25)
fsa_75[indf] = np.nanpercentile(slope_fsa,75)
fsa_min[indf] = np.nanmin(slope_fsa)
fsa_max[indf] = np.nanmax(slope_fsa)
unc_crdl[indf] = np.nanstd(crdl)/np.sqrt(nnn)
med_crdl[indf] = np.nanmedian(crdl)
crdl_25[indf] = np.nanpercentile(crdl,25)
crdl_75[indf] = np.nanpercentile(crdl,75)
crdl_min[indf] = np.nanmin(crdl)
crdl_max[indf] = np.nanmax(crdl)
unc_frdl[indf] = np.nanstd(frdl)/np.sqrt(nnn)
med_frdl[indf] = np.nanmedian(frdl)
frdl_25[indf] = np.nanpercentile(frdl,25)
frdl_75[indf] = np.nanpercentile(frdl,75)
frdl_min[indf] = np.nanmin(frdl)
frdl_max[indf] = np.nanmax(frdl)
header_txt = ('mdiam;udiam;diam25;diam75;diam_min;diam_max;' +
'mdepth;udepth;depth25;depth75;depth_min;depth_max;' +
'mh;uh;h25;h75;hmin;hmax;' +
'mhr;uhr;hr25;hr75;hrmin;hrmax;' +
'm_mcw;u_mcw;mcw_25;mcw_75;mcw_min;mcw_max;' +
'm_ucw;u_ucw;ucw_25;ucw_75;ucw_min;ucw_max;' +
'mcse;ucse;cse_25;cse_75;cse_min;cse_max;'
'mrupcw;urupcw;rupcw_25;rupcw_75;rupcw_min;rupcw_max;'
'mrufrc;urufrc;rufrc_25;rufrc_75;rufrc_min;rufrc_max;'
'mlrs;ulrs;lrs_25;lrs_75;lrs_min;lrs_max;'
'murs;uurs;urs_25;urs_75;urs_min;urs_max;'
'mfsa;ufsa;fsa_25;fsa_75;fsa_min;fsa_max;'
'mcrdl;ucrdl;crdl_25;crdl_75;crdl_min;crdl_max;'
'mfrdl;ufrdl;frdl_25;frdl_75;frdl_min;frdl_max')
except:
unc_cse[indf] = np.nan
med_cse[indf] = np.nan
cse_25[indf] = np.nan
cse_75[indf] = np.nan
cse_min[indf] = np.nan
cse_max[indf] = np.nan
unc_mcw[indf] = np.nan
med_mcw[indf] = np.nan
mcw_25[indf] = np.nan
mcw_75[indf] = np.nan
mcw_min[indf] = np.nan
mcw_max[indf] = np.nan
unc_ucw[indf] = np.nan
med_ucw[indf] = np.nan
ucw_25[indf] = np.nan
ucw_75[indf] = np.nan
ucw_min[indf] = np.nan
ucw_max[indf] = np.nan
unc_h[indf] = np.nan
med_h[indf] = np.nan
h_25[indf] = np.nan
h_75[indf] = np.nan
h_min[indf] = np.nan
h_max[indf] = np.nan
unc_hr[indf] = np.nan
med_hr[indf] = np.nan
hr_25[indf] = np.nan
hr_75[indf] = np.nan
hr_min[indf] = np.nan
hr_max[indf] = np.nan
depthf[indf] = np.nan
med_depth[indf] = np.nan
depth_25[indf] = np.nan
depth_75[indf] = np.nan
depth_min[indf] = np.nan
depth_max[indf] = np.nan
diamf[indf] = np.nan
med_diam[indf] = np.nan
diam_25[indf] = np.nan
diam_75[indf] = np.nan
diam_min[indf] = np.nan
diam_max[indf] = np.nan
##
unc_rupcw[indf] = np.nan
med_rupcw[indf] = np.nan
rupcw_25[indf] = np.nan
rupcw_75[indf] = np.nan
rupcw_min[indf] = np.nan
rupcw_max[indf] = np.nan
unc_rufrc[indf] = np.nan
med_rufrc[indf] = np.nan
rufrc_25[indf] = np.nan
rufrc_75[indf] = np.nan
rufrc_min[indf] = np.nan
rufrc_max[indf] = np.nan
unc_lrs[indf] = np.nan
med_lrs[indf] = np.nan
lrs_25[indf] = np.nan
lrs_75[indf] = np.nan
lrs_min[indf] = np.nan
lrs_max[indf] = np.nan
unc_urs[indf] = np.nan
med_urs[indf] = np.nan
urs_25[indf] = np.nan
urs_75[indf] = np.nan
urs_min[indf] = np.nan
urs_max[indf] = np.nan
unc_fsa[indf] = np.nan
med_fsa[indf] = np.nan
fsa_25[indf] = np.nan
fsa_75[indf] = np.nan
fsa_min[indf] = np.nan
fsa_max[indf] = np.nan
unc_crdl[indf] = np.nan
med_crdl[indf] = np.nan
crdl_25[indf] = np.nan
crdl_75[indf] = np.nan
crdl_min[indf] = np.nan
crdl_max[indf] = np.nan
unc_frdl[indf] = np.nan
med_frdl[indf] = np.nan
frdl_25[indf] = np.nan
frdl_75[indf] = np.nan
frdl_min[indf] = np.nan
frdl_max[indf] = np.nan
header_txt = ('mdiam;udiam;diam25;diam75;diam_min;diam_max;' +
'mdepth;udepth;depth25;depth75;depth_min;depth_max;' +
'mh;uh;h25;h75;hmin;hmax;' +
'mhr;uhr;hr25;hr75;hrmin;hrmax;' +
'm_mcw;u_mcw;mcw_25;mcw_75;mcw_min;mcw_max;' +
'm_ucw;u_ucw;ucw_25;ucw_75;ucw_min;ucw_max;' +
'mcse;ucse;cse_25;cse_75;cse_min;cse_max;'
'mrupcw;urupcw;rupcw_25;rupcw_75;rupcw_min;rupcw_max;'
'mrufrc;urufrc;rufrc_25;rufrc_75;rufrc_min;rufrc_max;'
'mlrs;ulrs;lrs_25;lrs_75;lrs_min;lrs_max;'
'murs;uurs;urs_25;urs_75;urs_min;urs_max;'
'mfsa;ufsa;fsa_25;fsa_75;fsa_min;fsa_max;'
'mcrdl;ucrdl;crdl_25;crdl_75;crdl_min;crdl_max;'
'mfrdl;ufrdl;frdl_25;frdl_75;frdl_min;frdl_max')
#if it fails cross sections does not work
# txt name
name_crater_f = filename + '_res.txt'
arc = np.column_stack((med_diam[indf], diamf[indf], diam_25[indf], diam_75[indf], diam_min[indf], diam_max[indf],
med_depth[indf], depthf[indf], depth_25[indf], depth_75[indf], depth_min[indf], depth_max[indf],
med_h[indf], unc_h[indf], h_25[indf], h_75[indf], h_min[indf], h_max[indf],
med_hr[indf], unc_hr[indf], hr_25[indf], hr_75[indf], hr_min[indf], hr_max[indf],
med_mcw[indf], unc_mcw[indf], mcw_25[indf], mcw_75[indf], mcw_min[indf], mcw_max[indf],
med_ucw[indf], unc_ucw[indf], ucw_25[indf], ucw_75[indf], ucw_min[indf], ucw_max[indf],
med_cse[indf], unc_cse[indf], cse_25[indf], cse_75[indf], cse_min[indf], cse_max[indf],
med_rupcw[indf], unc_rupcw[indf], rupcw_25[indf], rupcw_75[indf], rupcw_min[indf], rupcw_max[indf],
med_rufrc[indf], unc_rufrc[indf], rufrc_25[indf], rufrc_75[indf], rufrc_min[indf], rufrc_max[indf],
med_lrs[indf], unc_lrs[indf], lrs_25[indf], lrs_75[indf], lrs_min[indf], lrs_max[indf],
med_urs[indf], unc_urs[indf], urs_25[indf], urs_75[indf], urs_min[indf], urs_max[indf],
med_fsa[indf], unc_fsa[indf], fsa_25[indf], fsa_75[indf], fsa_min[indf], fsa_max[indf],
med_crdl[indf], unc_crdl[indf], crdl_25[indf], crdl_75[indf], crdl_min[indf], crdl_max[indf],
med_frdl[indf], unc_frdl[indf], frdl_25[indf], frdl_75[indf], frdl_min[indf], frdl_max[indf]))
np.savetxt(pathdata + name_crater_f, arc, delimiter = ";", header=header_txt,fmt='%10.5f', comments='#')
# added part about saving cross sections. I still need to test if this good or not
name_crater_crossX = filename + '_cross_sectionsX.txt'
name_crater_crossY = filename + '_cross_sectionsY.txt'
name_crater_crossZ = filename + '_cross_sectionsZ.txt'
# in order to save it I have to transform the dictionnary into an array
# I need to get the maximum length
maxlength_crossSections = 0
for xx_tmp, x_tmp in crossSections.items():
if len(x_tmp) > maxlength_crossSections:
maxlength_crossSections = len(x_tmp)
else:
None
number_crossSections = len(crossSections)
# create array
array_crossSections = np.empty((maxlength_crossSections, number_crossSections))
array_crossSections[:] = np.nan
arrayY_crossSections = np.empty((maxlength_crossSections, number_crossSections))
arrayY_crossSections[:] = np.nan
arrayX_crossSections = np.empty((maxlength_crossSections, number_crossSections))
arrayX_crossSections[:] = np.nan
# fill array with values
for xx_tmp, x_tmp in crossSections.items():
maxlength_crossSections = len(x_tmp)
array_crossSections[:maxlength_crossSections,xx_tmp] = x_tmp
arrayY_crossSections[:maxlength_crossSections,xx_tmp] = YSections[xx_tmp]
arrayX_crossSections[:maxlength_crossSections,xx_tmp] = XSections[xx_tmp]
np.savetxt(pathdata + name_crater_crossX, arrayX_crossSections, delimiter = ";",fmt='%10.5f', comments='#')
np.savetxt(pathdata + name_crater_crossY, arrayY_crossSections, delimiter = ";",fmt='%10.5f', comments='#')
np.savetxt(pathdata + name_crater_crossZ, array_crossSections, delimiter = ";",fmt='%10.5f', comments='#')
return (med_diam, diamf, diam_25, diam_75, diam_min, diam_max,
med_depth, depthf, depth_25, depth_75, depth_min, depth_max,
med_h, unc_h, h_25, h_75, h_min, h_max,
med_hr, unc_hr, hr_25, hr_75, hr_min, hr_max,
med_mcw, unc_mcw, mcw_25, mcw_75, mcw_min, mcw_max,
med_ucw, unc_ucw, ucw_25, ucw_75, ucw_min, ucw_max,
med_cse, unc_cse, cse_25, cse_75, cse_min, cse_max,
med_rupcw, unc_rupcw, rupcw_25, rupcw_75, rupcw_min, rupcw_max,
med_rufrc, unc_rufrc, rufrc_25, rufrc_75, rufrc_min, rufrc_max,
med_lrs, unc_lrs, lrs_25, lrs_75, lrs_min, lrs_max,
med_urs, unc_urs, urs_25, urs_75, urs_min, urs_max,
med_fsa, unc_fsa, fsa_25, fsa_75, fsa_min, fsa_max,
med_crdl, unc_crdl, crdl_25, crdl_75, crdl_min, crdl_max,
med_frdl, unc_frdl, frdl_25, frdl_75, frdl_min, frdl_max)
def run1(path, filenameXY, filenamecrater, pathplot, pathdata):
'''
filenameXY = 'data.txt'
filenamecrater = 'crater_id.txt'
'''
# the ASCII file (converted from the raster in ArcGIS) is loaded
os.chdir(path)
filenames = glob.glob(
"CPcrater*.asc") #need to fix that if coldspots are included?
filenames.sort(key=wk.tokenize)
dataXYD = np.loadtxt(filenameXY, delimiter=";", comments="#")
xcrater = dataXYD[:, 0]
ycrater = dataXYD[:, 1]
diam0 = dataXYD[:, 2]
diam0 = diam0 * 1000.
r0 = diam0 / 2.
#it takes way too much time this way
#xarcout = np.array([])
#yarcout = np.array([])
#if it is the last step
for indf, filename in enumerate(filenames):
print(indf)
# should be good this way
data = np.loadtxt(path + filename, skiprows=6)
data2 = np.rot90(data)
data3 = np.rot90(data2)
data4 = np.rot90(data3)
del data
del data2
del data3
data = copy.deepcopy(data4 *
0.50) #multiplying factor of the elevation
del data4
#correspond to orgin in the lower-left corner and data[x,y]
# with x = ncols, y = nrows !!! important
# the number of columns, rows and extent of the raster is read
(ncols, nrows, xllcorner, yllcorner, cellsize,
NODATA_value) = wk.readheader(path, filename)
# if not the same number of rows or columns
if ncols > nrows:
data = data[:nrows, :nrows]
ncols = nrows
elif nrows > ncols:
data = data[:ncols, :ncols]
nrows = ncols
else:
pass
#print diam
# x and y axes are created (x and y have the same size so it is not a big deal
# if we mix up here)
x = np.arange(0, ((ncols) * cellsize), cellsize)
y = np.arange(0, ((nrows) * cellsize), cellsize)
# x and y at the center of the cell
xe = np.arange(cellsize / 2., ((ncols) * cellsize), cellsize)
ye = np.arange(cellsize / 2., ((nrows) * cellsize), cellsize)
if len(x) > len(xe):
x = x[:len(xe)]
y = y[:len(xe)]
# create my own matrices for the plotting with pcolor or pcolormesh
xc = np.zeros_like(data)
yc = np.zeros_like(data)
xce = np.zeros_like(data)
yce = np.zeros_like(data)
for i in range(ncols):
xc[:, i] = x
xce[:, i] = xe
for i in range(nrows):
yc[i, :] = y
yce[i, :] = ye
# get the cell the closest to the center of the crater (this part has to be changed as the proj is changed)
#__, ncenterx = wk.find_nearest(xllcorner+xe, xcrater[indf]) #correspond to columns and x
# because yllcorner is from the top left now
#__, ncentery = wk.find_nearest(yllcorner+ye,ycrater[indf]) #correspond to rows and y
# middle of the dtm
ncentery = int(nrows /
2) #Python 3 770/2 is not directly integer anymore
ncenterx = int(nrows / 2)
# Average height at 1R, 2R, 3R, 3.5R and 4R from the centre of the crater
# (need to be replace with the right centery and x)
#other places where ncenterx and ncentery is required
x1, y1 = wk.xy_circle(1.0 * r0[indf], xe[ncenterx],
ye[ncentery]) # xe and ye are equals
#x2, y2 = wk.xy_circle(2.0*r0[indf], xe[ncenterx], ye[ncentery])
#x3, y3 = wk.xy_circle(3.0*r0[indf], xe[ncenterx], ye[ncentery])
#x35, y35 = wk.xy_circle(3.5*r0[indf], xe[ncenterx], ye[ncentery])
#x4, y4 = wk.xy_circle(4.0*r0[indf], xe[ncenterx], ye[ncentery])
# First detrending (a plane is fitted through the elevations taken at circles
# 2.0 and 3.0R from the center of the crater )
stduse = True #use standard deviation removal
ndata = wk.detrending(xc, yc, r0[indf], cellsize, ncenterx, ncentery,
data, stduse)
# txt name
name_crater_txt = filename.split('.asc')[0] + 'XY.txt'
#and if XY is not found in a certain path
if os.path.isfile(pathdata + name_crater_txt):
pass
else:
# the Maximum elevation are selected a first time
first_run = True
mingrade = 0.05
minclust = 0.05
slen = 0.1
(col_coord_ME, row_coord_ME, col_cells_ME, row_cells_ME, elev_ME,
prof_ME) = (wk.rim(xc, yc, ncenterx, ncentery, ndata, r0[indf],
cellsize, slen, minclust, mingrade, first_run))
# we get rid of the nan-values for the detrending
ixnan = np.where(np.isnan(col_coord_ME) == False)
# Maximum elevations without NaNs
coldetrend = col_coord_ME[ixnan]
rowdetrend = row_coord_ME[ixnan]
elev4detrend = elev_ME[ixnan]
# the second detrending through the selected maximum elevation points are done
stduse = False
Z2 = wk.linear3Ddetrending(coldetrend, rowdetrend, elev4detrend,
xc, yc, stduse)
# the detrended plane is substracted to the DEM
ndata2 = ndata - Z2
# the first detrending has ben run so we change now the first_run flag to false
first_run = False
# second run of the function rim
(col_coord_ME, row_coord_ME, col_cells_ME, row_cells_ME, elev_ME,
prof_ME, col_coord_LE, row_coord_LE, col_cells_LE, row_cells_LE,
elev_LE, prof_LE, col_coord_BS, row_coord_BS, col_cells_BS,
row_cells_BS, elev_BS,
prof_BS) = (wk.rim(xc, yc, ncenterx, ncentery, ndata2, r0[indf],
cellsize, slen, minclust, mingrade, first_run))
# takes only nonnan values from *BS data and merge it to local elevation data
ixfinite = np.where(np.isfinite(col_coord_BS) == True)
col_coord_BS = col_coord_BS[ixfinite]
row_coord_BS = row_coord_BS[ixfinite]
col_cells_BS = col_cells_BS[ixfinite]
row_cells_BS = row_cells_BS[ixfinite]
elev_BS = elev_BS[ixfinite]
prof_BS = prof_BS[ixfinite]
use_break_in_slope = False
if use_break_in_slope:
# concatenate local maxima and slope change!
colint = np.concatenate((col_coord_LE, col_coord_BS))
rowint = np.concatenate((row_coord_LE, row_coord_BS))
colmap = np.concatenate((col_cells_LE, col_cells_BS))
rowmap = np.concatenate((row_cells_LE, row_cells_BS))
elevint = np.concatenate((elev_LE, elev_BS))
profint = np.concatenate((prof_LE, prof_BS))
else:
# or using only local maxima
colint = col_coord_LE
rowint = row_coord_LE
colmap = col_cells_LE
rowmap = row_cells_LE
elevint = elev_LE
profint = prof_LE
# Maximum allowed radial discontinuity Drad (I should convert these values in cells)
Drad = 0.1 * r0[indf]
# Distance of interest (searching distance)
Dint = 0.05 * r0[indf]
# Maximum angular discontinuity (avoid unnecessary large gap angle in the
# data)
angle = 2.0 #(in degrees)
stangle = [0, 45, 90, 135, 180, 225, 270, 315]
contloop = True
siftRedundant = True
kpstitch = False
OptRims, Omegas, gap, maxradf = wk.rim_composite(
col_coord_ME, row_coord_ME, col_cells_ME, row_cells_ME,
elev_ME, prof_ME, colint, rowint, colmap, rowmap, xc, yc,
elevint, profint, ncenterx, ncentery, angle, stangle, Drad,
Dint, contloop, siftRedundant, kpstitch)
# we want to export the data as X, Y, Z
# find the best fit
for i in range(np.shape(OptRims)[0]):
a = np.mean(maxradf[i]) + (0.5 * Omegas[i])
if i == 0:
b = a
c = i
else:
if a < b:
b = a
c = i
# I should get rid of the 0,0 counts and should not append it takes way too much time
# temporary name
name_crater_png = filename.split('.asc')[0] + 'XY.png'
#xarcout = np.append(xarcout, np.array(OptRims[c][0,:]) + xllcorner)
#yarcout = np.append(yarcout, np.array(OptRims[c][1,:]) + yllcorner)
plt.ioff()
plt.figure()
plt.pcolormesh(xc, yc, ndata2)
plt.title(filename.split('.asc')[0], fontsize=16)
plt.colorbar()
xarcgis = np.array(OptRims[c][0, :])
xarcgis = xarcgis[np.nonzero(xarcgis)]
yarcgis = np.array(OptRims[c][1, :])
yarcgis = yarcgis[np.nonzero(yarcgis)]
zarcgis = np.array(OptRims[c][2, :])
zarcgis = zarcgis[np.nonzero(yarcgis)]
profgis = np.array(OptRims[c][3, :])
profgis = profgis[np.nonzero(yarcgis)]
flaggis = np.array(OptRims[c][4, :])
flaggis = flaggis[np.nonzero(yarcgis)]
#xarcgis = np.array(OptRims[c][0,:]) + xllcorner
#xarcgis = xarcgis[np.nonzero(xarcgis - xllcorner)]
#yarcgis = np.array(OptRims[c][1,:]) + yllcorner
#yarcgis = yarcgis[np.nonzero(yarcgis - yllcorner)]
#zarcgis = np.array(OptRims[c][2,:])
#zarcgis = zarcgis[np.nonzero(yarcgis - yllcorner)]
#profgis = np.array(OptRims[c][3,:])
#profgis = profgis[np.nonzero(yarcgis - yllcorner)]
#flaggis = np.array(OptRims[c][4,:])
#flaggis = flaggis[np.nonzero(yarcgis - yllcorner)]
#print np.shape(OptRims)
plt.plot(xarcgis, yarcgis, "ko")
plt.plot(x1, y1, "b")
# save plot
plt.savefig(pathplot + name_crater_png, dpi=300)
plt.close()
#export the data to a .csv file with ";"
#xarcgis = np.array(OptRims[c][0,:]) + xllcorner
#yarcgis = np.array(OptRims[c][1,:]) + yllcorner
#zarcgis = np.array(OptRims[c][2,:])
header_txt = 'X;Y;Z;prof;flag'
arcout = np.column_stack(
(xarcgis, yarcgis, zarcgis, profgis, flaggis))
np.savetxt(pathdata + name_crater_txt,
arcout,
delimiter=";",
header=header_txt,
fmt='%10.5f',
comments='')
def run2(path, filenameXY, filenamecrater, pathdata):
'''
filenameXY = 'data.txt'
filenamecrater = 'crater_id.txt'
'''
os.chdir(path)
filenames = glob.glob("CPcrater*.asc")
filenames.sort(key=wk.tokenize)
crater_id = np.genfromtxt(filenamecrater, skip_header=1, dtype=None)
dataXYD = np.loadtxt(filenameXY, delimiter=";", comments="#")
xcrater = dataXYD[:, 0]
ycrater = dataXYD[:, 1]
diam0 = dataXYD[:, 2]
diam0 = diam0 * 1000.
r0 = diam0 / 2.
unc_cse = np.ones(len(crater_id))
med_cse = np.ones(len(crater_id))
cse_25 = np.ones(len(crater_id))
cse_75 = np.ones(len(crater_id))
cse_min = np.ones(len(crater_id))
cse_max = np.ones(len(crater_id))
unc_mcw = np.ones(len(crater_id))
med_mcw = np.ones(len(crater_id))
mcw_25 = np.ones(len(crater_id))
mcw_75 = np.ones(len(crater_id))
mcw_min = np.ones(len(crater_id))
mcw_max = np.ones(len(crater_id))
unc_ucw = np.ones(len(crater_id))
med_ucw = np.ones(len(crater_id))
ucw_25 = np.ones(len(crater_id))
ucw_75 = np.ones(len(crater_id))
ucw_min = np.ones(len(crater_id))
ucw_max = np.ones(len(crater_id))
unc_h = np.ones(len(crater_id))
med_h = np.ones(len(crater_id))
h_25 = np.ones(len(crater_id))
h_75 = np.ones(len(crater_id))
h_min = np.ones(len(crater_id))
h_max = np.ones(len(crater_id))
med_depth = np.ones(len(crater_id))
depthf = np.ones(len(crater_id))
depth_25 = np.ones(len(crater_id))
depth_75 = np.ones(len(crater_id))
depth_min = np.ones(len(crater_id))
depth_max = np.ones(len(crater_id))
diamf = np.ones(len(crater_id))
med_diam = np.ones(len(crater_id))
diam_25 = np.ones(len(crater_id))
diam_75 = np.ones(len(crater_id))
diam_min = np.ones(len(crater_id))
diam_max = np.ones(len(crater_id))
for indf, filename in enumerate(filenames):
print(indf)
# load data
data = np.loadtxt(path + filename, skiprows=6)
data2 = np.rot90(data)
data3 = np.rot90(data2)
data4 = np.rot90(data3)
del data
del data2
del data3
data = copy.deepcopy(data4 * 0.50) # multiplying by scaling factor
del data4
# txt name
name_crater_txt = filename.split('.asc')[0] + 'XY.txt'
# the number of columns, rows and extent of the raster is read
(ncols, nrows, xllcorner, yllcorner, cellsize,
NODATA_value) = wk.readheader(path, filename)
# if not the same number of rows or columns
if ncols > nrows:
data = data[:nrows, :nrows]
ncols = nrows
elif nrows > ncols:
data = data[:ncols, :ncols]
nrows = ncols
else:
pass
# define a few variables
x = np.arange(0, ((ncols) * cellsize), cellsize)
y = np.arange(0, ((nrows) * cellsize), cellsize)
# x and y at the center of the cell
xe = np.arange(cellsize / 2., ((ncols) * cellsize), cellsize)
ye = np.arange(cellsize / 2., ((nrows) * cellsize), cellsize)
if len(x) > len(xe):
x = x[:len(xe)]
y = y[:len(xe)]
# create my own matrices for the plotting with pcolor or pcolormesh
xc = np.zeros_like(data)
yc = np.zeros_like(data)
xce = np.zeros_like(data)
yce = np.zeros_like(data)
for i in range(ncols):
xc[:, i] = x
xce[:, i] = xe
for i in range(nrows):
yc[i, :] = y
yce[i, :] = ye
#load the data
datagis = loadgis(pathdata, name_crater_txt)
xarcgis = datagis[:, 0]
yarcgis = datagis[:, 1]
zarcgis = datagis[:, 2]
profgis = datagis[:, 3]
flaggis = datagis[:, 4]
#get rid of outliers
#Q1 = np.percentile(zarcgis,25.)
#Q3 = np.percentile(zarcgis,75.)
#IQR = Q3 - Q1
#outliers = (zarcgis < (Q1 - 1.5 * IQR)) | (zarcgis > (Q3 + 1.5 * IQR))
#not_outliers = np.where(outliers == False)
x_not_outliers = xarcgis
y_not_outliers = yarcgis
z_not_outliers = zarcgis
prof_not_outliers = profgis
# old center
# get the cell the closest to the center of the crater
#__, ncenterx_old = wk.find_nearest(xllcorner+xe, xcrater[indf]) #correspond to columns and x
# because yllcorner is from the top left now
#__, ncentery_old = wk.find_nearest(yllcorner+ye,ycrater[indf]) #correspond to rows and y
ncenterx_old = nrows / 2
ncentery_old = nrows / 2
#fit a circle
xnewcenter, ynewcenter, rnew, residu = wk.leastsq_circle(
x_not_outliers, y_not_outliers)
# get the cell the closest to the center of the crater (new calculation)
__, ncenterx = wk.find_nearest(xe, xnewcenter)
__, ncentery = wk.find_nearest(ye, ynewcenter)
#other places where ncenterx and ncentery is required (new calculation)
#x1, y1 = wk.xy_circle(1.0*rnew, xe[ncentery], ye[ncenterx])
# First detrending (a plane is fitted through the elevations taken at circles
# with new center of the circle
try:
stduse = True
ndata = wk.detrending(xc, yc, rnew, cellsize, ncenterx_old,
ncentery_old, data, stduse) #detrending
# with old data otherwise sometimes the dem area is not large enough
# the second detrending through the selected maximum elevation points are done
stduse = False
Z2 = wk.linear3Ddetrending(x_not_outliers, y_not_outliers,
z_not_outliers, xc, yc, stduse)
# the detrended plane is substracted to the DEM (with the newly detected)
ndata2 = ndata - Z2
# I should actually rerun the rim detection things one more time
# as the centre of the crater has been moved (not the same profiles anymore)
#need to put some indices there (maybe, they will have different sizes)
(R_upcw, R_ufrc, cse, slope_mcw, slope_ucw, slope_fsa, slope_lrs,
slope_urs, h, depth, diam, nnn, prf, crossSections, YSections,
XSections) = (wk.calculation(xc, yc, x_not_outliers,
y_not_outliers, z_not_outliers,
prof_not_outliers, rnew, ndata2,
cellsize, ncenterx, ncentery,
xllcorner, yllcorner))
# crossSections has been added
# the problem is that there are several values for profile = 0? for some reasons?
# I don't understand that, I guess it must be profiles where....
# np.where(prf != 0) to get rid of it but it would be nice if the problem would be detected
#just need to make the calculation, we have the final circle
# need to find the closest values that does not overlap
unc_cse[indf] = np.nanstd(cse) / np.sqrt(nnn)
med_cse[indf] = np.nanmedian(cse)
cse_25[indf] = np.nanpercentile(cse, 25)
cse_75[indf] = np.nanpercentile(cse, 75)
cse_min[indf] = np.nanmin(cse)
cse_max[indf] = np.nanmax(cse)
unc_mcw[indf] = np.nanstd(slope_mcw) / np.sqrt(nnn)
med_mcw[indf] = np.nanmedian(slope_mcw)
mcw_25[indf] = np.nanpercentile(slope_mcw, 25)
mcw_75[indf] = np.nanpercentile(slope_mcw, 75)
mcw_min[indf] = np.nanmin(slope_mcw)
mcw_max[indf] = np.nanmax(slope_mcw)
unc_ucw[indf] = np.nanstd(slope_ucw) / np.sqrt(nnn)
med_ucw[indf] = np.nanmedian(slope_ucw)
ucw_25[indf] = np.nanpercentile(slope_ucw, 25)
ucw_75[indf] = np.nanpercentile(slope_ucw, 75)
ucw_min[indf] = np.nanmin(slope_ucw)
ucw_max[indf] = np.nanmax(slope_ucw)
unc_h[indf] = np.nanstd(h) / np.sqrt(nnn)
med_h[indf] = np.nanmedian(h)
h_25[indf] = np.nanpercentile(h, 25)
h_75[indf] = np.nanpercentile(h, 75)
h_min[indf] = np.nanmin(h)
h_max[indf] = np.nanmax(h)
depthf[indf] = np.nanstd(depth) / np.sqrt(nnn)
med_depth[indf] = np.nanmedian(depth)
depth_25[indf] = np.nanpercentile(depth, 25)
depth_75[indf] = np.nanpercentile(depth, 75)
depth_min[indf] = np.nanmin(depth)
depth_max[indf] = np.nanmax(depth)
diamf[indf] = np.nanstd(diam) / np.sqrt(nnn)
med_diam[indf] = np.nanmedian(diam)
diam_25[indf] = np.nanpercentile(diam, 25)
diam_75[indf] = np.nanpercentile(diam, 75)
diam_min[indf] = np.nanmin(diam)
diam_max[indf] = np.nanmax(diam)
header_txt = (
'mdiam;udiam;diam25;diam75;diam_min;diam_max;' +
'mdepth;udepth;depth25;depth75;depth_min;depth_max;' +
'mh;uh;h25;h75;hmin;hmax;' +
'm_mcw;u_mcw;mcw_25;mcw_75;mcw_min;mcw_max;' +
'm_ucw;u_ucw;ucw_25;ucw_75;ucw_min;ucw_max;' +
'mcse;ucse;cse_25;cse_75;cse_min;cse_max;')
except:
unc_cse[indf] = np.nan
med_cse[indf] = np.nan
cse_25[indf] = np.nan
cse_75[indf] = np.nan
cse_min[indf] = np.nan
cse_max[indf] = np.nan
unc_mcw[indf] = np.nan
med_mcw[indf] = np.nan
mcw_25[indf] = np.nan
mcw_75[indf] = np.nan
mcw_min[indf] = np.nan
mcw_max[indf] = np.nan
unc_ucw[indf] = np.nan
med_ucw[indf] = np.nan
ucw_25[indf] = np.nan
ucw_75[indf] = np.nan
ucw_min[indf] = np.nan
ucw_max[indf] = np.nan
unc_h[indf] = np.nan
med_h[indf] = np.nan
h_25[indf] = np.nan
h_75[indf] = np.nan
h_min[indf] = np.nan
h_max[indf] = np.nan
depthf[indf] = np.nan
med_depth[indf] = np.nan
depth_25[indf] = np.nan
depth_75[indf] = np.nan
depth_min[indf] = np.nan
depth_max[indf] = np.nan
diamf[indf] = np.nan
med_diam[indf] = np.nan
diam_25[indf] = np.nan
diam_75[indf] = np.nan
diam_min[indf] = np.nan
diam_max[indf] = np.nan
header_txt = (
'mdiam;udiam;diam25;diam75;diam_min;diam_max;' +
'mdepth;udepth;depth25;depth75;depth_min;depth_max;' +
'mh;uh;h25;h75;hmin;hmax;' +
'm_mcw;u_mcw;mcw_25;mcw_75;mcw_min;mcw_max;' +
'm_ucw;u_ucw;ucw_25;ucw_75;ucw_min;ucw_max;' +
'mcse;ucse;cse_25;cse_75;cse_min;cse_max;')
# txt name
name_crater_f = filename.split('.asc')[0] + '_res.txt'
arc = np.column_stack(
(med_diam[indf], diamf[indf], diam_25[indf], diam_75[indf],
diam_min[indf], diam_max[indf], med_depth[indf], depthf[indf],
depth_25[indf], depth_75[indf], depth_min[indf], depth_max[indf],
med_h[indf], unc_h[indf], h_25[indf], h_75[indf], h_min[indf],
h_max[indf], med_mcw[indf], unc_mcw[indf], mcw_25[indf],
mcw_75[indf], mcw_min[indf], mcw_max[indf], med_ucw[indf],
unc_ucw[indf], ucw_25[indf], ucw_75[indf], ucw_min[indf],
ucw_max[indf], med_cse[indf], unc_cse[indf], cse_25[indf],
cse_75[indf], cse_min[indf], cse_max[indf]))
np.savetxt(pathdata + name_crater_f,
arc,
delimiter=";",
header=header_txt,
fmt='%10.5f',
comments='#')
# added part about saving cross sections. I still need to test if this good or not
name_crater_cross = filename.split('.asc')[0] + '_cross_sections.txt'
# in order to save it I have to transform the dictionnary into an array
# I need to get the maximum length
maxlength_crossSections = 0
for xx_tmp, x_tmp in crossSections.items():
if len(x_tmp) > maxlength_crossSections:
maxlength_crossSections = len(x_tmp)
else:
None
number_crossSections = len(crossSections)
# create array
array_crossSections = np.empty(
(maxlength_crossSections, number_crossSections))
array_crossSections[:] = np.nan
arrayY_crossSections = np.empty(
(maxlength_crossSections, number_crossSections))
arrayY_crossSections[:] = np.nan
arrayX_crossSections = np.empty(
(maxlength_crossSections, number_crossSections))
arrayX_crossSections[:] = np.nan
# fill array with values
for xx_tmp, x_tmp in crossSections.items():
maxlength_crossSections = len(x_tmp)
array_crossSections[:maxlength_crossSections, xx_tmp] = x_tmp
arrayY_crossSections[:maxlength_crossSections,
xx_tmp] = YSections[xx_tmp]
arrayX_crossSections[:maxlength_crossSections,
xx_tmp] = XSections[xx_tmp]
datacross = np.column_stack(
(arrayY_crossSections, arrayX_crossSections, array_crossSections))
np.savetxt(pathdata + name_crater_cross,
datacross,
delimiter=";",
fmt='%10.5f',
comments='#')
return (med_diam, diamf, diam_25, diam_75, diam_min, diam_max, med_depth,
depthf, depth_25, depth_75, depth_min, depth_max, med_h, unc_h,
h_25, h_75, h_min, h_max, med_mcw, unc_mcw, mcw_25, mcw_75,
mcw_min, mcw_max, med_ucw, unc_ucw, ucw_25, ucw_75, ucw_min,
ucw_max, med_cse, unc_cse, cse_25, cse_75, cse_min, cse_max)