-
Notifications
You must be signed in to change notification settings - Fork 0
/
diag.trajectory.drifters.colloc.extrapolated.histogram.jl
317 lines (297 loc) · 18.3 KB
/
diag.trajectory.drifters.colloc.extrapolated.histogram.jl
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
#=
= Loop through the available collocations and grid the corresponding forward and backward
= extrapolations relative to the actual (uninterpolated) values. Note that BEF refers to
= a forecast extrapolation using analysis data from before the target value and AFT refers
= to a revcast extrapolation using data from afterward. Relative to the values at the
= extrapolation target time (TOTN), both local (binwise) and global regressions of the
= two separate extrapolations (from before and after) are also saved. The global
= regression is then used to adjust the data and the same files are stored. This defines
= an extrapolation bias correction. Extrapolation variance (relative to the uninterpolated
= values) is also calculated and stored, both for all collocations and as a function of
= current speed - RD May, November 2016, January, March 2017.
=#
using My
const ODAT = 1 # identify indecies of the input data:
const OLAT = 2 # date/lat/lon/obs on the collocation grid
const OLON = 3
const OCUR = 4
const MISS = -9999.0 # generic missing value
const CUTOFF = 200
if (argc = length(ARGS)) != 2
print("\nUsage: jjj $(basename(@__FILE__)) buoydata_1993_2014_drogON.asc.nonmdt.locate_2.0_extra.ucur.got2000_obs.comb v2.0_global_025_deg_total_15m\n\n")
exit(1)
end
shift = -1
ARGS[2] == "v2.0_global_025_deg_ekman_15m" && (shift = 0)
ARGS[2] == "v2.0_global_025_deg_ekman_hs" && (shift = 3)
ARGS[2] == "v2.0_global_025_deg_geostrophic" && (shift = 6)
ARGS[2] == "v2.0_global_025_deg_total_15m" && (shift = 9)
ARGS[2] == "v2.0_global_025_deg_total_hs" && (shift = 12)
if shift > -1
const TOTB = 5 + shift # identify the three analysis indecies
const TOTN = 6 + shift
const TOTA = 7 + shift
else
print("\nERROR: $(ARGS[2]) is not recognized\n\n")
exit(-1)
end
step = 0.02 ; bound = collect(-4.0:step:4.0)
gridbefnow = zeros(length(bound), length(bound)) ; conbefnow = zeros(length(bound))
gridaftnow = zeros(length(bound), length(bound)) ; conaftnow = zeros(length(bound))
gridbefobs = zeros(length(bound), length(bound)) ; conbefobs = zeros(length(bound)) ; conobsbef = zeros(length(bound))
gridnowobs = zeros(length(bound), length(bound)) ; connowobs = zeros(length(bound)) ; conobsnow = zeros(length(bound))
gridaftobs = zeros(length(bound), length(bound)) ; conaftobs = zeros(length(bound)) ; conobsaft = zeros(length(bound))
gridbefaft = zeros(length(bound), length(bound)) ; conbefaft = zeros(length(bound)) ; conaftbef = zeros(length(bound))
regressbef = Array(Float64, 0)
regressnow = Array(Float64, 0)
regressaft = Array(Float64, 0)
regressobs = Array(Float64, 0)
binbef = fill(Array(Float64, 0), length(bound))
binnow = fill(Array(Float64, 0), length(bound))
binaft = fill(Array(Float64, 0), length(bound))
cutbef = Array(Float64, length(bound), CUTOFF)
cutnow = Array(Float64, length(bound), CUTOFF)
cutaft = Array(Float64, length(bound), CUTOFF)
if contains(ARGS[1], "wcur") # substitute speed for current component
fila = replace(ARGS[1], "wcur", "ucur")
filb = replace(ARGS[1], "wcur", "vcur")
fpa = My.ouvre(fila, "r") ; tinea = readlines(fpa) ; close(fpa) ; tinuma = length(tinea)
fpb = My.ouvre(filb, "r") ; tineb = readlines(fpb) ; close(fpb) ; tinumb = length(tineb)
tinuma == tinumb || (print("\nERROR: tinuma $tinuma != $tinumb tinumb\n\n") ; exit(-1))
for a = 1:tinuma
vala = float(split(tinea[a]))
valb = float(split(tineb[a]))
out = tinea[a][1:30]
for b = 4:19
valc = (vala[b]^2 + valb[b]^2)^0.5
out *= @sprintf(" %9.3f", valc)
end
tinea[a] = out
end
else
fpa = My.ouvre(ARGS[1], "r") ; tinea = readlines(fpa) ; close(fpa) ; tinuma = length(tinea)
end
for a = 1:tinuma # grid the collocations and save values
vals = float(split(tinea[a])) # for a regression versus TOTN and OCUR
if vals[TOTB] > -333 && vals[TOTB] < 333 &&
vals[TOTN] > -333 && vals[TOTN] < 333 &&
vals[TOTA] > -333 && vals[TOTA] < 333 &&
vals[OCUR] > -333 && vals[OCUR] < 333
delbef, indbef = findmin(abs(bound - vals[TOTB])) ; bound[indbef] > vals[TOTB] && indbef > 1 && (indbef -= 1)
delnow, indnow = findmin(abs(bound - vals[TOTN])) ; bound[indnow] > vals[TOTN] && indnow > 1 && (indnow -= 1)
delaft, indaft = findmin(abs(bound - vals[TOTA])) ; bound[indaft] > vals[TOTA] && indaft > 1 && (indaft -= 1)
delobs, indobs = findmin(abs(bound - vals[OCUR])) ; bound[indobs] > vals[OCUR] && indobs > 1 && (indobs -= 1)
gridbefnow[indbef,indnow] += 1 ; conbefnow[indnow] += vals[TOTB]
gridaftnow[indaft,indnow] += 1 ; conaftnow[indnow] += vals[TOTA]
gridbefobs[indbef,indobs] += 1 ; conbefobs[indobs] += vals[TOTB] ; conobsbef[indbef] += vals[OCUR]
gridnowobs[indnow,indobs] += 1 ; connowobs[indobs] += vals[TOTN] ; conobsnow[indnow] += vals[OCUR]
gridaftobs[indaft,indobs] += 1 ; conaftobs[indobs] += vals[TOTA] ; conobsaft[indaft] += vals[OCUR]
gridbefaft[indbef,indaft] += 1 ; conbefaft[indaft] += vals[TOTB] ; conaftbef[indbef] += vals[TOTA]
push!(regressbef, vals[TOTB])
push!(regressnow, vals[TOTN])
push!(regressaft, vals[TOTA])
push!(regressobs, vals[OCUR])
tmp = copy(binbef[indnow]) ; push!(tmp, vals[TOTB]) ; binbef[indnow] = tmp
tmp = copy(binnow[indnow]) ; push!(tmp, vals[TOTN]) ; binnow[indnow] = tmp
tmp = copy(binaft[indnow]) ; push!(tmp, vals[TOTA]) ; binaft[indnow] = tmp
end
end
valsb = Array(Float64, tinuma) # also grid for the extrapolation variance
valsn = Array(Float64, tinuma) # (versus TOTN) using all values in each bin
valsa = Array(Float64, tinuma) # (above) and for the nearest CUTOFF values
dista = Array(Float64, tinuma) # (here, but only save nearest CUTOFFs below)
linuma = tinuma < CUTOFF ? tinuma : CUTOFF
for a = 1:tinuma vals = float(split(tinea[a])) ; valsb[a] = vals[TOTB] ; valsn[a] = vals[TOTN] ; valsa[a] = vals[TOTA] end
for (z, ranz) in enumerate(bound)
for a = 1:tinuma dista[a] = abs(ranz - valsn[a]) end ; lima = sort(dista)[linuma] ; b = 1
for a = 1:tinuma
if dista[a] <= lima && b <= linuma
cutbef[z,b] = valsb[a]
cutnow[z,b] = valsn[a]
cutaft[z,b] = valsa[a]
b += 1
end
end
end
fname = ARGS[1] * "." * ARGS[2] * ".extra.dat"
fpa = My.ouvre(fname, "w") # and save the grids
for (a, vala) in enumerate(bound)
for (b, valb) in enumerate(bound)
@printf(fpa, "%15.8f %15.8f %15.8f %15.8f %15.8f %15.8f\n", gridbefnow[b,a], gridaftnow[b,a], gridbefobs[b,a], gridaftobs[b,a], gridbefaft[b,a], gridnowobs[b,a])
end
end
close(fpa)
sumbefnow = sum(gridbefnow, 1) # as well as the corresponding count and sum
sumaftnow = sum(gridaftnow, 1) # of extrapolation values in each TOTN interval
sumbefobs = sum(gridbefobs, 1) ; sumobsbef = sum(gridbefobs, 2)
sumnowobs = sum(gridnowobs, 1) ; sumobsnow = sum(gridnowobs, 2)
sumaftobs = sum(gridaftobs, 1) ; sumobsaft = sum(gridaftobs, 2)
sumbefaft = sum(gridbefaft, 1) ; sumaftbef = sum(gridbefaft, 2)
fname = ARGS[1] * "." * ARGS[2] * ".extra.sum"
fpa = My.ouvre(fname, "w")
for (a, vala) in enumerate(bound)
@printf(fpa, "%22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %18.11e %18.11e %22d %8.4f\n",
sumbefnow[a], conbefnow[a], sumaftnow[a], conaftnow[a],
sumbefobs[a], conbefobs[a], sumaftobs[a], conaftobs[a],
sumobsbef[a], conobsbef[a], sumobsaft[a], conobsaft[a],
sumbefaft[a], conbefaft[a], sumaftbef[a], conaftbef[a],
sumnowobs[a], connowobs[a], sumobsnow[a], conobsnow[a],
cov(vec(cutbef[a,:]), vec(cutnow[a,:])), cov(vec(cutaft[a,:]), vec(cutnow[a,:])), length(binbef[a]), bound[a])
end
close(fpa)
fname = ARGS[1] * "." * ARGS[2] * ".extra.reg" # and finally save the regression coefficients
fpa = My.ouvre(fname, "w")
(intbefnow, slobefnow) = linreg(regressnow, regressbef)
(intaftnow, sloaftnow) = linreg(regressnow, regressaft)
(intbefobs, slobefobs) = linreg(regressobs, regressbef) ; (intobsbef, sloobsbef) = linreg(regressbef, regressobs)
(intnowobs, slonowobs) = linreg(regressobs, regressnow) ; (intobsnow, sloobsnow) = linreg(regressnow, regressobs)
(intaftobs, sloaftobs) = linreg(regressobs, regressaft) ; (intobsaft, sloobsaft) = linreg(regressaft, regressobs)
(intbefaft, slobefaft) = linreg(regressaft, regressbef) ; (intaftbef, sloaftbef) = linreg(regressbef, regressaft)
covbefnow = cov(regressnow, regressbef)
covaftnow = cov(regressnow, regressaft)
covbefobs = cov(regressobs, regressbef) ; covobsbef = cov(regressbef, regressobs)
covnowobs = cov(regressobs, regressnow) ; covobsnow = cov(regressnow, regressobs)
covaftobs = cov(regressobs, regressaft) ; covobsaft = cov(regressaft, regressobs)
covbefaft = cov(regressaft, regressbef) ; covaftbef = cov(regressbef, regressaft)
@printf(fpa, "%33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f\n",
intbefnow, slobefnow, intaftnow, sloaftnow,
intbefobs, slobefobs, intaftobs, sloaftobs,
intobsbef, sloobsbef, intobsaft, sloobsaft,
intbefaft, slobefaft, intaftbef, sloaftbef,
intnowobs, slonowobs, intobsnow, sloobsnow,
covbefnow, covaftnow,
covbefobs, covaftobs,
covobsbef, covobsaft,
covbefaft, covaftbef,
covnowobs, covobsnow)
close(fpa)
gridbefnow = zeros(length(bound), length(bound)) ; conbefnow = zeros(length(bound))
gridaftnow = zeros(length(bound), length(bound)) ; conaftnow = zeros(length(bound))
gridbefobs = zeros(length(bound), length(bound)) ; conbefobs = zeros(length(bound)) ; conobsbef = zeros(length(bound))
gridnowobs = zeros(length(bound), length(bound)) ; connowobs = zeros(length(bound)) ; conobsnow = zeros(length(bound))
gridaftobs = zeros(length(bound), length(bound)) ; conaftobs = zeros(length(bound)) ; conobsaft = zeros(length(bound))
gridbefaft = zeros(length(bound), length(bound)) ; conbefaft = zeros(length(bound)) ; conaftbef = zeros(length(bound))
regressbef = Array(Float64, 0)
regressnow = Array(Float64, 0)
regressaft = Array(Float64, 0) # reinitialize the variables for calibration
regressobs = Array(Float64, 0)
binbef = fill(Array(Float64, 0), length(bound))
binnow = fill(Array(Float64, 0), length(bound))
binaft = fill(Array(Float64, 0), length(bound))
cutbef = Array(Float64, length(bound), CUTOFF)
cutnow = Array(Float64, length(bound), CUTOFF)
cutaft = Array(Float64, length(bound), CUTOFF)
for a = 1:tinuma # now calibrate the bef and aft collocations
vals = float(split(tinea[a])) # using global regressions above and regrid
if vals[TOTB] > -333 && vals[TOTB] < 333 && # (but omit calibration of now versus obs)
vals[TOTN] > -333 && vals[TOTN] < 333 &&
vals[TOTA] > -333 && vals[TOTA] < 333 &&
vals[OCUR] > -333 && vals[OCUR] < 333
tmpb = (vals[TOTB] - intbefnow) /slobefnow
tmpn = vals[TOTN]
tmpa = (vals[TOTA] - intaftnow) /sloaftnow
tmpo = vals[OCUR]
delbef, indbef = findmin(abs(bound - tmpb)) ; bound[indbef] > tmpb && indbef > 1 && (indbef -= 1)
delnow, indnow = findmin(abs(bound - tmpn)) ; bound[indnow] > tmpn && indnow > 1 && (indnow -= 1)
delaft, indaft = findmin(abs(bound - tmpa)) ; bound[indaft] > tmpa && indaft > 1 && (indaft -= 1)
delobs, indobs = findmin(abs(bound - tmpo)) ; bound[indobs] > tmpo && indobs > 1 && (indobs -= 1)
gridbefnow[indbef,indnow] += 1 ; conbefnow[indnow] += tmpb
gridaftnow[indaft,indnow] += 1 ; conaftnow[indnow] += tmpa
gridbefobs[indbef,indobs] += 1 ; conbefobs[indobs] += tmpb ; conobsbef[indbef] += tmpo
gridnowobs[indnow,indobs] += 1 ; connowobs[indobs] += tmpn ; conobsnow[indnow] += tmpo
gridaftobs[indaft,indobs] += 1 ; conaftobs[indobs] += tmpa ; conobsaft[indaft] += tmpo
gridbefaft[indbef,indaft] += 1 ; conbefaft[indaft] += tmpb ; conaftbef[indbef] += tmpa
push!(regressbef, tmpb)
push!(regressnow, tmpn)
push!(regressaft, tmpa)
push!(regressobs, tmpo)
tmp = copy(binbef[indnow]) ; push!(tmp, tmpb) ; binbef[indnow] = tmp
tmp = copy(binnow[indnow]) ; push!(tmp, tmpn) ; binnow[indnow] = tmp
tmp = copy(binaft[indnow]) ; push!(tmp, tmpa) ; binaft[indnow] = tmp
end
end
valsb = Array(Float64, tinuma) # also grid for the extrapolation variance
valsn = Array(Float64, tinuma) # (versus TOTN) using all values in each bin
valsa = Array(Float64, tinuma) # (above) and for the nearest CUTOFF values
dista = Array(Float64, tinuma) # (here, but only save nearest CUTOFFs below)
linuma = tinuma < CUTOFF ? tinuma : CUTOFF
for a = 1:tinuma vals = float(split(tinea[a])) ; valsb[a] = (vals[TOTB] - intbefnow) /slobefnow ; valsn[a] = vals[TOTN] ; valsa[a] = (vals[TOTA] - intaftnow) /sloaftnow end
for (z, ranz) in enumerate(bound)
for a = 1:tinuma dista[a] = abs(ranz - valsn[a]) end ; lima = sort(dista)[linuma] ; b = 1
for a = 1:tinuma
if dista[a] <= lima && b <= linuma
cutbef[z,b] = valsb[a]
cutnow[z,b] = valsn[a]
cutaft[z,b] = valsa[a]
b += 1
end
end
end
fname = ARGS[1] * "." * ARGS[2] * ".extra.dau"
fpa = My.ouvre(fname, "w") # and save the grids
for (a, vala) in enumerate(bound)
for (b, valb) in enumerate(bound)
@printf(fpa, "%15.8f %15.8f %15.8f %15.8f %15.8f %15.8f\n", gridbefnow[b,a], gridaftnow[b,a], gridbefobs[b,a], gridaftobs[b,a], gridbefaft[b,a], gridnowobs[b,a])
end
end
close(fpa)
sumbefnow = sum(gridbefnow, 1) # as well as the corresponding count and sum
sumaftnow = sum(gridaftnow, 1) # of extrapolation values in each TOTN interval
sumbefobs = sum(gridbefobs, 1) ; sumobsbef = sum(gridbefobs, 2)
sumnowobs = sum(gridnowobs, 1) ; sumobsnow = sum(gridnowobs, 2)
sumaftobs = sum(gridaftobs, 1) ; sumobsaft = sum(gridaftobs, 2)
sumbefaft = sum(gridbefaft, 1) ; sumaftbef = sum(gridbefaft, 2)
fname = ARGS[1] * "." * ARGS[2] * ".extra.sun"
fpa = My.ouvre(fname, "w")
for (a, vala) in enumerate(bound)
@printf(fpa, "%22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %22.0f %33.11f %18.11e %18.11e %22d %8.4f\n",
sumbefnow[a], conbefnow[a], sumaftnow[a], conaftnow[a],
sumbefobs[a], conbefobs[a], sumaftobs[a], conaftobs[a],
sumobsbef[a], conobsbef[a], sumobsaft[a], conobsaft[a],
sumbefaft[a], conbefaft[a], sumaftbef[a], conaftbef[a],
sumnowobs[a], connowobs[a], sumobsnow[a], conobsnow[a],
cov(vec(cutbef[a,:]), vec(cutnow[a,:])), cov(vec(cutaft[a,:]), vec(cutnow[a,:])), length(binbef[a]), bound[a])
end
close(fpa)
fname = ARGS[1] * "." * ARGS[2] * ".extra.reh" # and finally save the regression coefficients
fpa = My.ouvre(fname, "w")
(intbefnow, slobefnow) = linreg(regressnow, regressbef)
(intaftnow, sloaftnow) = linreg(regressnow, regressaft)
(intbefobs, slobefobs) = linreg(regressobs, regressbef) ; (intobsbef, sloobsbef) = linreg(regressbef, regressobs)
(intnowobs, slonowobs) = linreg(regressobs, regressnow) ; (intobsnow, sloobsnow) = linreg(regressnow, regressobs)
(intaftobs, sloaftobs) = linreg(regressobs, regressaft) ; (intobsaft, sloobsaft) = linreg(regressaft, regressobs)
(intbefaft, slobefaft) = linreg(regressaft, regressbef) ; (intaftbef, sloaftbef) = linreg(regressbef, regressaft)
covbefnow = cov(regressnow, regressbef)
covaftnow = cov(regressnow, regressaft)
covbefobs = cov(regressobs, regressbef) ; covobsbef = cov(regressbef, regressobs)
covnowobs = cov(regressobs, regressnow) ; covobsnow = cov(regressnow, regressobs)
covaftobs = cov(regressobs, regressaft) ; covobsaft = cov(regressaft, regressobs)
covbefaft = cov(regressaft, regressbef) ; covaftbef = cov(regressbef, regressaft)
@printf(fpa, "%33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f %33.11f\n",
intbefnow, slobefnow, intaftnow, sloaftnow,
intbefobs, slobefobs, intaftobs, sloaftobs,
intobsbef, sloobsbef, intobsaft, sloobsaft,
intbefaft, slobefaft, intaftbef, sloaftbef,
intnowobs, slonowobs, intobsnow, sloobsnow,
covbefnow, covaftnow,
covbefobs, covaftobs,
covobsbef, covobsaft,
covbefaft, covaftbef,
covnowobs, covobsnow)
close(fpa)
exit(0)
#=
buoydata_1993_2014_drogON.asc.nonmdt.locate_2.0_extra.ucur.got2000_obs.comb
dat, lat, lon, obs, bef1, now1, aft1, bef2, now2, aft2, bef3, now3, aft3, bef4, now4, aft4, bef5, now5, aft5
count = 0
for a = 1:length(bound)
b = length(binbef[a])
c = b >= 10 ? cov(binbef[a], binnow[a]) : 0
d = b >= 10 ? cov(binaft[a], binnow[a]) : 0
f = cov(vec(cutbef[a,:]), vec(cutnow[a,:]))
g = cov(vec(cutaft[a,:]), vec(cutnow[a,:]))
@printf("%8.4f %8d %18.14f %18.14f %18.14e %18.14e\n", bound[a], b, c, d, f, g)
count += b
end
exit(0)
=#