[180] | 1 | def min (field,axis=None): |
---|
| 2 | import numpy as np |
---|
[349] | 3 | if field is None: return None |
---|
[398] | 4 | if type(field).__name__=='MaskedArray': |
---|
| 5 | field.set_fill_value(np.NaN) |
---|
| 6 | return np.ma.array(field).min(axis=axis) |
---|
| 7 | elif (np.isnan(np.sum(field)) and (type(field).__name__ not in 'MaskedArray')): |
---|
| 8 | return np.ma.masked_invalid(field).min(axis=axis) |
---|
[349] | 9 | else: return np.array(field).min(axis=axis) |
---|
[180] | 10 | |
---|
| 11 | def max (field,axis=None): |
---|
| 12 | import numpy as np |
---|
[345] | 13 | if field is None: return None |
---|
[398] | 14 | if type(field).__name__=='MaskedArray': |
---|
| 15 | field.set_fill_value(np.NaN) |
---|
| 16 | return np.ma.array(field).max(axis=axis) |
---|
| 17 | elif (np.isnan(np.sum(field)) and (type(field).__name__ not in 'MaskedArray')): |
---|
| 18 | return np.ma.masked_invalid(field).max(axis=axis) |
---|
[345] | 19 | else: return np.array(field).max(axis=axis) |
---|
[180] | 20 | |
---|
| 21 | def mean (field,axis=None): |
---|
| 22 | import numpy as np |
---|
[349] | 23 | if field is None: return None |
---|
[391] | 24 | else: |
---|
| 25 | if type(field).__name__=='MaskedArray': |
---|
| 26 | field.set_fill_value(np.NaN) |
---|
| 27 | zout=np.ma.array(field).mean(axis=axis) |
---|
[427] | 28 | if axis is not None: |
---|
| 29 | zout.set_fill_value(np.NaN) |
---|
| 30 | return zout.filled() |
---|
| 31 | else:return zout |
---|
[398] | 32 | elif (np.isnan(np.sum(field)) and (type(field).__name__ not in 'MaskedArray')): |
---|
| 33 | zout=np.ma.masked_invalid(field).mean(axis=axis) |
---|
[427] | 34 | if axis is not None: |
---|
| 35 | zout.set_fill_value([np.NaN]) |
---|
| 36 | return zout.filled() |
---|
| 37 | else:return zout |
---|
[391] | 38 | else: |
---|
| 39 | return np.array(field).mean(axis=axis) |
---|
[180] | 40 | |
---|
[525] | 41 | def sum (field,axis=None): |
---|
| 42 | import numpy as np |
---|
| 43 | if field is None: return None |
---|
| 44 | else: |
---|
| 45 | if type(field).__name__=='MaskedArray': |
---|
| 46 | field.set_fill_value(np.NaN) |
---|
| 47 | zout=np.ma.array(field).sum(axis=axis) |
---|
| 48 | if axis is not None: |
---|
| 49 | zout.set_fill_value(np.NaN) |
---|
| 50 | return zout.filled() |
---|
| 51 | else:return zout |
---|
| 52 | elif (np.isnan(np.sum(field)) and (type(field).__name__ not in 'MaskedArray')): |
---|
| 53 | zout=np.ma.masked_invalid(field).sum(axis=axis) |
---|
| 54 | if axis is not None: |
---|
| 55 | zout.set_fill_value([np.NaN]) |
---|
| 56 | return zout.filled() |
---|
| 57 | else:return zout |
---|
| 58 | else: |
---|
| 59 | return np.array(field).sum(axis=axis) |
---|
[647] | 60 | |
---|
| 61 | def getmask (field): |
---|
| 62 | import numpy as np |
---|
| 63 | if field is None: return None |
---|
| 64 | if type(field).__name__=='MaskedArray': |
---|
| 65 | return np.ma.getmask(field) |
---|
| 66 | else: |
---|
| 67 | return np.isnan(field) |
---|
| 68 | |
---|
[525] | 69 | |
---|
[180] | 70 | def deg (): |
---|
| 71 | return u'\u00b0' |
---|
| 72 | |
---|
[310] | 73 | def writeascii ( tab, filename ): |
---|
| 74 | mydata = tab |
---|
| 75 | myfile = open(filename, 'w') |
---|
| 76 | for line in mydata: |
---|
[647] | 77 | zeline = str(line) |
---|
| 78 | zeline = zeline.replace('[','') |
---|
| 79 | zeline = zeline.replace(']','') |
---|
| 80 | myfile.write(zeline + '\n') |
---|
[310] | 81 | myfile.close() |
---|
| 82 | return |
---|
| 83 | |
---|
[388] | 84 | |
---|
| 85 | # A.C. routine to compute saturation temperature |
---|
[391] | 86 | # Be Carefull, when asking for tsat-t, this routine outputs a masked array. |
---|
| 87 | # To be correctly handled, this call to tsat must be done before the call to |
---|
| 88 | # reduce_field, which handles correctly masked array with the new mean() function. |
---|
[388] | 89 | def get_tsat(pressure,temp=None,zlon=None,zlat=None,zalt=None,ztime=None): |
---|
| 90 | import math as mt |
---|
| 91 | import numpy as np |
---|
| 92 | acond=3.2403751E-04 |
---|
| 93 | bcond=7.3383721E-03 |
---|
| 94 | # if temp is not in input, the routine simply outputs the vertical profile |
---|
| 95 | # of Tsat |
---|
| 96 | if temp is None: |
---|
| 97 | # Identify dimensions in temperature field |
---|
| 98 | output=np.zeros(np.array(pressure).shape) |
---|
| 99 | if len(np.array(pressure).shape) is 1: |
---|
| 100 | #pressure field is a 1d column, (i.e. the altitude coordinate) |
---|
| 101 | #temperature has to have a z-axis |
---|
| 102 | i=0 |
---|
| 103 | for pp in pressure: |
---|
| 104 | output[i]=1./(bcond-acond*mt.log(.0095*pp)) |
---|
| 105 | i=i+1 |
---|
| 106 | else: |
---|
| 107 | #pressure field is a field present in the file. Unhandled |
---|
| 108 | #by this routine for now, which only loads unique variables. |
---|
| 109 | print "3D pressure field not handled for now, exiting in tsat" |
---|
| 110 | print "Use a vertical pressure coordinate if you want to compute Tsat" |
---|
| 111 | exit() |
---|
| 112 | # if temp is in input, the routine computes Tsat-T by detecting where the |
---|
| 113 | # vertical axis is in temp |
---|
| 114 | else: |
---|
| 115 | output=np.zeros(np.array(temp).shape) |
---|
| 116 | vardim=get_dim(zlon,zlat,zalt,ztime,temp) |
---|
| 117 | if 'altitude' not in vardim.keys(): |
---|
| 118 | print 'no altitude coordinate in temperature field for Tsat computation' |
---|
| 119 | exit() |
---|
| 120 | zdim=vardim['altitude'] |
---|
| 121 | ndim=len(np.array(temp).shape) |
---|
| 122 | print '--- in tsat(). vardim,zdim,ndim: ',vardim,zdim,ndim |
---|
| 123 | i=0 |
---|
| 124 | for pp in pressure: |
---|
| 125 | if ndim is 1: |
---|
| 126 | output[i]=1./(bcond-acond*mt.log(.0095*pp))-temp[i] |
---|
| 127 | elif ndim is 2: |
---|
| 128 | if zdim is 0: |
---|
| 129 | output[i,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[i,:] |
---|
| 130 | elif zdim is 1: |
---|
| 131 | output[:,i]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,i] |
---|
| 132 | else: |
---|
| 133 | print "stop in get_tsat: zdim: ",zdim |
---|
| 134 | exit() |
---|
| 135 | elif ndim is 3: |
---|
| 136 | if zdim is 0: |
---|
| 137 | output[i,:,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[i,:,:] |
---|
| 138 | elif zdim is 1: |
---|
| 139 | output[:,i,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,i,:] |
---|
| 140 | elif zdim is 2: |
---|
| 141 | output[:,:,i]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,:,i] |
---|
| 142 | else: |
---|
| 143 | print "stop in get_tsat: zdim: ",zdim |
---|
| 144 | exit() |
---|
| 145 | elif ndim is 4: |
---|
| 146 | if zdim is 0: |
---|
| 147 | output[i,:,:,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[i,:,:,:] |
---|
| 148 | elif zdim is 1: |
---|
| 149 | output[:,i,:,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,i,:,:] |
---|
| 150 | elif zdim is 2: |
---|
| 151 | output[:,:,i,:]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,:,i,:] |
---|
| 152 | elif zdim is 3: |
---|
| 153 | output[:,:,:,i]=1./(bcond-acond*mt.log(.0095*pp))-temp[:,:,:,i] |
---|
| 154 | else: |
---|
| 155 | print "stop in get_tsat: zdim: ", zdim |
---|
| 156 | exit() |
---|
| 157 | else: |
---|
| 158 | print "stop in get_tsat: ndim: ",ndim |
---|
| 159 | exit() |
---|
| 160 | i=i+1 |
---|
[391] | 161 | m=np.ma.masked_invalid(temp,copy=False) |
---|
| 162 | zoutput=np.ma.array(output,mask=m.mask,fill_value=np.NaN) |
---|
| 163 | return zoutput |
---|
[388] | 164 | |
---|
| 165 | # A.C. Dirty routine to determine where are the axis of a variable |
---|
| 166 | def get_dim(zlon,zlat,zalt,ztime,zvar): |
---|
| 167 | import numpy as np |
---|
| 168 | nx,ny,nz,nt=0,0,0,0 |
---|
| 169 | if zlon is not None: |
---|
| 170 | nx=len(zlon) |
---|
| 171 | if zlat is not None: |
---|
| 172 | ny=len(zlat) |
---|
| 173 | if zalt is not None: |
---|
| 174 | nz=len(zalt) |
---|
| 175 | if ztime is not None: |
---|
| 176 | nt=len(ztime) |
---|
| 177 | zdims={} |
---|
| 178 | zdims['longitude']=nx |
---|
| 179 | zdims['latitude']=ny |
---|
| 180 | zdims['altitude']=nz |
---|
| 181 | zdims['Time']=nt |
---|
| 182 | zvardim=np.array(zvar).shape |
---|
| 183 | ndim=len(zvardim) |
---|
| 184 | zzvardim=[[]]*ndim |
---|
| 185 | j=0 |
---|
| 186 | output={} |
---|
| 187 | for dim in zvardim: |
---|
| 188 | if dim not in zdims.values(): |
---|
| 189 | print "WARNING -----------------------------" |
---|
| 190 | print "Dimensions given to subroutine do not match variables dimensions :" |
---|
| 191 | exit() |
---|
| 192 | else: |
---|
| 193 | a=get_key(zdims,dim) |
---|
| 194 | if len(a) is not 1: |
---|
| 195 | if j is 0: ##this should solve most conflicts with Time |
---|
| 196 | zzvardim[j]=a[1] |
---|
| 197 | else: |
---|
| 198 | zzvardim[j]=a[0] |
---|
| 199 | else: |
---|
| 200 | zzvardim[j]=a[0] |
---|
| 201 | output[zzvardim[j]]=j |
---|
| 202 | j=j+1 |
---|
| 203 | return output |
---|
| 204 | |
---|
| 205 | # A.C. routine that gets keys from a dictionnary value |
---|
| 206 | def get_key(self, value): |
---|
| 207 | """find the key(s) as a list given a value""" |
---|
| 208 | return [item[0] for item in self.items() if item[1] == value] |
---|
| 209 | |
---|
[403] | 210 | # A.C. routine that gets the nearest value index of array and value |
---|
| 211 | def find_nearest(arr,value,axis=None,strict=False): |
---|
| 212 | import numpy as np |
---|
| 213 | # Special case when the value is nan |
---|
| 214 | if value*0 != 0: return np.NaN |
---|
| 215 | # Check that the value we search is inside the array for the strict mode |
---|
| 216 | if strict: |
---|
| 217 | min=arr.min() |
---|
| 218 | max=arr.max() |
---|
| 219 | if ((value > max) or (value < min)): return np.NaN |
---|
| 220 | |
---|
| 221 | if type(arr).__name__=='MaskedArray': |
---|
| 222 | mask=np.ma.getmask(arr) |
---|
| 223 | idx=np.ma.argmin(np.abs(arr-value),axis=axis) |
---|
| 224 | # Special case when there are only missing values on the axis |
---|
| 225 | if mask[idx]: |
---|
| 226 | idx=np.NaN |
---|
| 227 | else: |
---|
| 228 | idx=(np.abs(arr-value)).argmin(axis=axis) |
---|
| 229 | return idx |
---|
| 230 | |
---|
[792] | 231 | # Author: A.C. |
---|
[430] | 232 | def fig2data ( fig ): |
---|
| 233 | import numpy |
---|
| 234 | """ |
---|
| 235 | @brief Convert a Matplotlib figure to a 4D numpy array with RGBA channels and return it |
---|
| 236 | @param fig a matplotlib figure |
---|
| 237 | @return a numpy 3D array of RGBA values |
---|
| 238 | """ |
---|
| 239 | # draw the renderer |
---|
| 240 | fig.canvas.draw ( ) |
---|
| 241 | |
---|
| 242 | # Get the RGBA buffer from the figure |
---|
| 243 | w,h = fig.canvas.get_width_height() |
---|
| 244 | buf = numpy.fromstring ( fig.canvas.tostring_argb(), dtype=numpy.uint8 ) |
---|
| 245 | buf.shape = ( w, h,4 ) |
---|
| 246 | |
---|
| 247 | # canvas.tostring_argb give pixmap in ARGB mode. Roll the ALPHA channel to have it in RGBA mode |
---|
| 248 | buf = numpy.roll ( buf, 3, axis = 2 ) |
---|
| 249 | return buf |
---|
| 250 | |
---|
[792] | 251 | # Author: A.C. |
---|
[430] | 252 | def fig2img ( fig ): |
---|
| 253 | import Image |
---|
| 254 | import numpy |
---|
| 255 | """ |
---|
| 256 | @brief Convert a Matplotlib figure to a PIL Image in RGBA format and return it |
---|
| 257 | @param fig a matplotlib figure |
---|
| 258 | @return a Python Imaging Library ( PIL ) image |
---|
| 259 | """ |
---|
| 260 | # put the figure pixmap into a numpy array |
---|
| 261 | buf = fig2data ( fig ) |
---|
| 262 | w, h, d = buf.shape |
---|
| 263 | return Image.fromstring( "RGBA", ( w ,h ), buf.tostring( ) ) |
---|
[792] | 264 | |
---|
| 265 | # Author: A.C. |
---|
| 266 | # Convert a single layer image object (greyscale) to an array |
---|
| 267 | def image2array(im): |
---|
| 268 | import numpy as np |
---|
| 269 | if im.mode not in ("L", "F"): |
---|
| 270 | raise ValueError, ("can only convert single-layer images", im.mode) |
---|
| 271 | if im.mode == "L": |
---|
| 272 | a = np.fromstring(im.tostring(), np.uint8) |
---|
| 273 | else: |
---|
| 274 | a = np.fromstring(im.tostring(), np.float32) |
---|
| 275 | a.shape = im.size[1], im.size[0] |
---|
| 276 | return a |
---|
| 277 | |
---|
| 278 | # Author: A.C. |
---|
| 279 | # Convert a 2D array to a single layer image object (greyscale) |
---|
| 280 | def array2image(a): |
---|
| 281 | import numpy as np |
---|
| 282 | import Image |
---|
| 283 | if a.dtype == np.uint8: |
---|
| 284 | mode = "L" |
---|
| 285 | elif a.dtype == np.float32: |
---|
| 286 | mode = "F" |
---|
| 287 | else: |
---|
| 288 | raise ValueError, "unsupported image mode" |
---|
| 289 | return Image.fromstring(mode, (a.shape[1], a.shape[0]), a.tostring()) |
---|
| 290 | |
---|