-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathensembles.Rmd
353 lines (288 loc) · 13.2 KB
/
ensembles.Rmd
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
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
---
title: ""
output: rmarkdown::html_vignette
vignette: >
%\VignetteIndexEntry{Ensembles}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>"
)
options(tibble.print_min = 4L, tibble.print_max = 4L)
options(rmarkdown.html_vignette.check_title = FALSE)
```
# r2ogs5 ensembles
An ensemble refers to several simulations of the same type with different input parameters. Here a set of helper functions are presented in order to set up and run an ensemble.
1. [Prerequisites](#inst)
2. [Simulation setup](#setup)
3. [Preprocessing](#preproc)
4. [Ensemble run](#sim)
5. [Postprocessing](#post)
6. [Visualization](#viz)
## Prerequisites <a name="inst"></a>
Before being able to run any simulation, the package r2ogs needs to be installed
via
```{r, eval=FALSE}
devtools::install_git("https://gitlab.opengeosys.org/ogs5/r2ogs5")
# alternatively:
remotes::install_git("https://gitlab.opengeosys.org/ogs5/r2ogs5")
```
in R or
```{bash, eval=FALSE}
git clone https://gitlab.opengeosys.org/ogs5/r2ogs5.git
cd r2ogs5
R --slave -e 'devtools::install(".")'
```
from a terminal, requiring the package *devtools* being installed in the systems R library.
```{r}
library(r2ogs5)
```
```{r, echo= FALSE, eval=TRUE}
options(r2ogs5.default_ogs5_bin = search_ogs5_bin_path(ogs_exe = "ogs_fem", return_ogs5_bin = TRUE))
```
An appropriate *OGS5* executable has to be compiled or downloaded and the name
(here ogs_fem) and location of the OGS5 executable have to be set accordingly
via `options`, for example:
```{r, eval=FALSE}
options(r2ogs5.default_ogs5_bin = "/usr/local/bin/ogs_fem")
```
**Note:** In order to install the package with vignettes, e.g. by calling
```{r, eval=FALSE}
devtools::install_git("https://gitlab.opengeosys.org/ogs5/r2ogs5", build_vignettes = TRUE)
```
a basic ogs5 executable named "ogs_fem" must be available in the search path. For Linux, the executable should be in one of the directories found when entering `Sys.getenv("PATH")` in the R console. For Windows, it should be located in *"C:/Programme/OpenGeoSys/bin/"* or *"C:/Programs/OpenGeoSys/bin/"* under the current implementation.
## Simulation setup <a name="setup"></a>
Before setting up the ensemble, a base *ogs5* simulation has to be defined.
The input files will be read in from an already set up simulation, whose input files should be in the directory specified in `temp_dir`. The same
hydraulic transport model is utilized as in `vignette("r2ogs5")`, for the creation of the
input files used here please refer to this vignette.
```{r, results='hide'}
temp_dir <- "tmp"
ex1 <- create_ogs5(sim_name = "ex1", sim_id = 1L,
sim_path = temp_dir)
```
```{r, results='hide', eval = TRUE, echo=FALSE}
ex <- create_ogs5(sim_name = "ex1", sim_id = 1L,
sim_path = temp_dir)
# this chunk is not visible in the vignette but ensures that the files exist in the
# directory when the vignette is built.
ex <- input_add_pcs_bloc(x=ex, pcs_name = "waterflow", PCS_TYPE = "LIQUID_FLOW",
PRIMARY_VARIABLE = "PRESSURE1")
ex <- input_add_pcs_bloc(x=ex, pcs_name = "tracer",
PCS_TYPE = "MASS_TRANSPORT",
PRIMARY_VARIABLE = "Tracer1",
RELOAD = c("1", "1"))
ex <- input_add_bc_bloc(x=ex, bc_name = "pressure_out",
PRIMARY_VARIABLE = "PRESSURE1",
DIS_TYPE = "CONSTANT 9810",
PCS_TYPE = "LIQUID_FLOW",
GEO_TYPE = "POINT point1")
ex <- input_add_bc_bloc(x=ex, bc_name = "tracer_in",
PRIMARY_VARIABLE = "Tracer",
DIS_TYPE = "CONSTANT 1.19438",
PCS_TYPE = "MASS_TRANSPORT",
GEO_TYPE = "POINT point0",
TIM_TYPE = "CURVE 1")
ex <- input_add_ic_bloc(x = ex, ic_name = "pressure",
PCS_TYPE = "LIQUID_FLOW",
PRIMARY_VARIABLE = "PRESSURE1",
GEO_TYPE = "DOMAIN",
DIS_TYPE = "CONSTANT 9810")
ex <- input_add_ic_bloc(x = ex, ic_name = "tracer",
PCS_TYPE = "MASS_TRANSPORT",
PRIMARY_VARIABLE = "Tracer",
GEO_TYPE = "DOMAIN",
DIS_TYPE = "CONSTANT 0.0")
ex <- input_add_mcp_bloc(x = ex,
NAME = "Tracer",
MOBILE = 1)
ex <- input_add_mfp_bloc(x = ex, FLUID_NAME = "water",
FLUID_TYPE = "LIQUID",
DENSITY = "1 1.0E+3",
VISCOSITY = "1 1.0E-3")
ex <- input_add_mmp_bloc(x = ex, GEOMETRY_DIMENSION = "1",
GEOMETRY_AREA = "1.2",
POROSITY = "1 0.38",
PERMEABILITY_TENSOR = "ISOTROPIC 1.0E-8",
MASS_DISPERSION = "1 1.0",
TORTUOSITY = "1 1.0",
NAME = "gravel")
ex <- input_add_msp_bloc(x = ex, NAME = "gravel" ,
DENSITY = "1 1.8E+3")
ex <- input_add_num_bloc(x = ex, num_name = "waterflow",
PCS_TYPE = "LIQUID_FLOW",
LINEAR_SOLVER = "2 6 1.0E-14 100 1.0 100 4")
ex <- input_add_num_bloc(x = ex, num_name = "tracer",
PCS_TYPE = "MASS_TRANSPORT",
LINEAR_SOLVER = "2 6 1.0E-14 100 1.0 100 4")
ex <- input_add_out_bloc(x = ex, out_name = "waterflow",
PCS_TYPE = "LIQUID_FLOW",
NOD_VALUES = "PRESSURE1\n VELOCITY_X1",
GEO_TYPE = "DOMAIN",
DAT_TYPE = "PVD",
TIM_TYPE = "STEPS 10")
ex <- input_add_out_bloc(x = ex, out_name = "tracer",
PCS_TYPE = "MASS_TRANSPORT",
NOD_VALUES = "Tracer",
GEO_TYPE = "DOMAIN",
DAT_TYPE = "PVD",
TIM_TYPE = "STEPS 10")
ex <- input_add_out_bloc(x = ex, out_name = "tracer_tec",
PCS_TYPE = "MASS_TRANSPORT",
NOD_VALUES = "Tracer",
GEO_TYPE = "DOMAIN",
DAT_TYPE = "TECPLOT", TIM_TYPE = "STEPS 1")
ex <- input_add_st_bloc(x = ex, st_name = "waterflow",
PCS_TYPE = "LIQUID_FLOW",
PRIMARY_VARIABLE = "PRESSURE1",
GEO_TYPE = "POINT point0",
DIS_TYPE = "CONSTANT_NEUMANN 7.8611E-06")
ex <- input_add_tim_bloc(x = ex, tim_name = "waterflow",
PCS_TYPE = "LIQUID_FLOW",
TIME_START = "0",
TIME_END = "36000000",
TIME_STEPS = "229 3600")
ex <- input_add_tim_bloc(x = ex, tim_name = "tracer",
PCS_TYPE = "MASS_TRANSPORT",
TIME_START = "0",
TIME_END = "36000000",
TIME_STEPS = "229 3600")
ex <- input_add_rfd_bloc(x = ex, rfd_name = "tracer",
mkey = "CURVES",
data = tibble::tibble(
time = c(0, 3600, 3600.1, 720, 36000000),
conc = c(1,1,0,0,0)))
ex <- input_add_gli_points(x = ex,
ogs5_points = tibble::tibble(
x = c(0, 4.7),
y = c(0,0),
z = c(0,0),
name = c("point0", "point1")))
# ad msh bloc
mesh_lst <- create_structured_mesh_nodes_ele(lx = 4.7, nx = 94)
ex <- input_add_msh_bloc(x = ex, msh_name = "base_mesh",
NODES = mesh_lst$NODES,
ELEMENTS = mesh_lst$ELEMENTS)
ogs5_write_inputfiles(ex, type = "all")
```
```{r, results='hide'}
ex1 <- input_add_blocs_from_file(
ogs5_obj = ex1,
sim_basename = "ex1",
filename = "all",
file_dir = temp_dir # the directory where input files were written to in
# vignette("r2ogs5")
)
```
## Preprocessing <a name="preproc"></a>
Define a parameter table for the ensemble for two parameters that will be arranged on a grid. Here, every combination of 10 mass dispersion and 5 tracer input concentrations is a point on the grid thus resulting in 50 points/rows in the table. This can be easily extended to more parameters, but it has to be kept in mind that the number of points is the product of the parameter sequences every point corresponds to a simulation run.
```{r results = "markup"}
para_df <-
expand.grid(mass_dispersion = seq(0.1, 2.5, length.out = 3),
tracer_input_concentration = seq(0.1, 1, length.out = 3)) %>%
tibble::as_tibble()
head(para_df)
```
Next, an ensemble object of class *ens* is created, where all the information is stored and a path to the directory where ensemble files should be written.
```{r}
ens1 = create_ens(base_sim = ex1, parameter_tbl = para_df,
name = "tracersim",
path = temp_dir)
names(attributes(ens1))
```
In a simple loop, each parameter combination can be changed in the base simulation by overwriting a temporary *ogs5* object "tmp_ogs5" and then adding it to the ensemble with the helper function `ens_add_ogs5`.
```{r}
# ensemble_add_sims <- functoin(ensemble = list())
# add sims
for (i in 1:dim(attributes(ens1)$sim_plan)[1]) {
# define new ogs5 obj based on attributes(ens1)$base_sim
tmp_ogs5 <- attributes(ens1)$base_sim
# set parameters
tmp_ogs5$input$mmp$gravel$MASS_DISPERSION <- paste(
"1 ",
round(attributes(ens1)$sim_plan[i, "mass_dispersion"], 2)
)
tmp_ogs5$input$bc$tracer_in$DIS_TYPE <- paste(
"CONSTANT",
attributes(ens1)$sim_plan[i, "tracer_input_concentration"]
)
# add to ens1
ens1 <- ens_add_ogs5(x = ens1, ogs5_obj = tmp_ogs5)
}
```
With another helper function, a new folder "tracersim" will be created in "tmp" and input files will be written into corresponding numbered sub-folders in the specified ensemble path. The general folder structure is:
ensemble_path/
\ \ \ \ - ensemble_name/
\ \ \ \ \ \ \ \ - ensemble_name1
\ \ \ \ \ \ \ \ - ensemble_name2
\ \ \ \ \ \ \ \ - ...
```{r}
ens_write_inputfiles(ens1)
```
## Simulate <a name="sim"></a>
The simulations are executed with the wrapper function `ens_run` that calls the function `ogs5_run`. Documentation is available under `?ogs5_run`.
```{r, results='hide', warning=FALSE, message=FALSE}
ens_run(ens1,
ogs_exe = NULL, # default path will be used
log_output = TRUE,
wait = TRUE)
```
The output defined in the third output block is read in next.
```{r, results='hide', message=FALSE, warning=FALSE}
ens1 <- ens_get_output(ens1, type = "specific", list("OUTPUT3"))
```
```{r}
ens1$sim1$output$OUTPUT3[[1]] %>% head
```
## Postprocessing <a name="post"></a>
For every simulation of the ensemble, there exists a *tibble* in `$output`. In order to use the plotting grammar from **ggplot2**, a single data table in long format is often useful. Therefore, the data from is stacked into a single *tibble* with the code below. A new column called "sim_name" keeps track of which results belong to which simulations.
```{r}
# create tibble of all outputted *.tec files
tracer_out_df <- tibble::tibble()
for (i in 1:length(ens1)){
tracer_out_df <-
tracer_out_df %>%
dplyr::bind_rows(
ens1[[i]]$output$OUTPUT3[[1]] %>%
dplyr::mutate(sim_name = attributes(ens1[[i]])$sim_name)
)
}
```
Next, the parameter table, for which simulations were run is matched with the respective simulation output by the column `sim_name`.
```{r}
tracer_out_df <-
tracer_out_df %>%
dplyr::filter(X == 4.7) %>%
dplyr::left_join(attributes(ens1)$sim_plan,
by= "sim_name") %>%
dplyr::mutate(TIME = TIME/(3600*24))
```
## Visualization <a name="viz"></a>
An overview of the effect on the simulation of the different parameters can be achieved by plotting, if there are not too many parameters. Here, the Tracer concentration at the last point (`x = 4.7`) is plotted over time as so-called breakthrough curves. It can be observed that both increasing tracer input concentration and result in slightly fatter tails of the curve.
```{r, eval=FALSE, fig.asp=0.5, fig.width=8}
g <- tracer_out_df %>%
dplyr::filter(X == 4.7) %>%
ggplot2::ggplot(ggplot2::aes(
x = TIME,
y = mass_dispersion,
height = Tracer,
group = mass_dispersion,
fill = Tracer
))+
ggridges::geom_density_ridges_gradient(stat = "identity",
rel_min_height = 0.01,
panel_scaling = TRUE)+
ggplot2::facet_grid(.~tracer_input_concentration)+
ggplot2::theme(
axis.text.x.bottom = ggplot2::element_text(angle = 30)
)+
ggplot2::labs(x = "time (days)",
subtitle = expression(input~c[Tracer]~(mg~L^{-1})),
y = expression(Dispersion~Length~(m)),
fill = expression(c[Tracer]~(mg~L^{-1}))
)
```
