-
Notifications
You must be signed in to change notification settings - Fork 0
/
tempclusters.Rmd
384 lines (274 loc) · 12.6 KB
/
tempclusters.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
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
---
title: "Temperature cluster analysis"
author: "Rosie"
date: "7/16/2020"
output:
pdf_document: default
html_document: default
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(lubridate)
library(tidyverse)
library(vegan)
library(sf)
library(RColorBrewer)
library(gridExtra)
library(scales)
library(ggmap)
temps = readRDS("Temp_filtered (1).rds")
```
## Cluster analysis
So, Larry was unceratain as to how to do cluster analysis when we "just have one variable". While it's true that we only have temperature to work with, if we really only just had one variable we could group stations with similar average temperatures and be done with it. However, we really do have a multivariate dataset because each station has different average temperatures at different times of year, and different variance in tempareture. We want to group stations with similar temperature regimes, rather than just similar averages.
Now, there are a number of different ways to describe the temperature regime, but I thought I would start by taking the average daily mean temperature over the past five years.
I used Cat's continuous data set rather than the descrete data set because it is a more balanced design where we will have a mean temperature for each day of the year and we know it was taken at the exact same spot.
First clean up the data and make sure we only have complete years.
```{r}
temps2 = mutate(temps, Year = year(Datetime)) %>%
group_by(Year, Station) %>%
summarize(numb = length(Temp)) %>%
filter(numb >7000, Year >= 2015)
temps3 = group_by(temps2, Station) %>%
filter(length(unique(Year)) >=4) %>%
ungroup()
tempssub = merge(temps, temps3)
```
```{r warning=FALSE, message=FALSE, fig.width= 12}
#first calculate the daily means
tempmean = tempssub %>%
filter(Date > as.Date("2014-1-1"), Station != "RYF") %>%
mutate(julian = yday(Date)) %>%
group_by(Station, julian) %>%
summarize(Temp = mean(Temp, na.rm = T))
#put it into wide format for the cluster analysis
tempwide = pivot_wider(tempmean, id_cols = c(Station),
names_from = julian, values_from = Temp)
row.names(tempwide) = tempwide$Station
#calculate distance and cluster
tempdist = dist(tempwide, "euclidean")
tempfit = hclust(tempdist, method = "ward.D")
plot(tempfit, main = "Clusters based on daily ave temp", cex = 0.6)
```
## Or we could use monthly means, mins, and maxes
I haven't sat down to compare the monthly mins, means, and maxes with the daily means, but a quick look seems like they are pretty similar
```{r warning=FALSE, message=FALSE, fig.width=12}
#monthly mean, min, and max per year
tempmo = temps %>%
filter(Date > as.Date("2014-1-1"), Station != "RYF") %>%
mutate(Month = month(Date), Year = year(Date)) %>%
group_by(Station, Month, Year) %>%
summarize(Temp = mean(Temp, na.rm = T), min = min(Temp), max = max(Temp))
#average over five years
tempmo2 = tempmo %>%
group_by(Station, Month) %>%
summarize(Temp = mean(Temp, na.rm = T), min = mean(min, na.rm = T), max = mean(max, na.rm = T))
#put it into wide format for the cluster analysis
tempmowide = pivot_wider(tempmo2, id_cols = c(Station),
names_from = Month, values_from = Temp)
tempmowide2 = pivot_wider(tempmo2, id_cols = c(Station),
names_from = Month, values_from = min)
tempmowide3 = pivot_wider(tempmo2, id_cols = c(Station),
names_from = Month, values_from = max)
tempmowide4 = cbind(tempmowide, tempmowide2[,-1], tempmowide3[,-1])
row.names(tempmowide4) = tempmowide4$Station
#calculate distance and cluster
tempdist3 = dist(tempmowide4, "euclidean")
tempfit3 = hclust(tempdist3, method = "ward.D")
plot(tempfit3, main = "Clusters based on monthly ave, min max", cex = 0.6)
```
I thought I'd look at a quick NMDS too, just for fun
```{r warning=FALSE, message=FALSE, echo=FALSE}
tempNMDS = metaMDS(tempdist3, trymax = 200)
plot(tempNMDS, type = "n")
text(tempNMDS, "sites", labels = tempmowide4$Station)
```
## CDEC stations
Now the question is whether teh clusters actually work out geographically. I'm going to have to spend some time with this map printed out and a highlighter to see how it lines up.
![CDEC stations included in analysis](CDEC_stations.jpg)
## Cutting down the trees
Now I can cut my trees into groups to see how they map out.
```{r}
cutday = as.data.frame(cutree(tempfit, k = c(2,4,6,12)))
cutday$Station = row.names(cutday)
```
## Plot
I trimmed the tree into different numbers of groups, so see what we get when we try to divide it into 4,6, or 12 regions. Personally, I like the 6 group version best.
```{r}
#read in shapefile of the delta
delta = read_sf("DeltaShapefile/hydro_delta_marsh.shp")
#read in shapefile of EDSM regions
edsm = read_sf("EDSM_Subregions/EDSM_Subregions_03302020.shp")
#reference system is UTMs
st_crs(edsm) = 32610
edsm = st_transform(edsm, 4326)
#add lat/longs for the stations
stas = read.csv("StationLatLongs.csv")
cutday = merge(stas, cutday, all.x =T)
names(cutday) = c("Station", "Latitude", "Longitude", "grps2", "grps4", "grps6", "grps12")
#turn it into a spatial object
stashap = st_as_sf(cutday,
coords = c("Longitude", "Latitude"),
crs = 4326)
save(stashap, file ="stations.rdata")
#Make plots of the stations divided into four groups, six groups
#or twelve groups based on teh cluster analysis.
gp4 = ggplot() +
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps4)))+
theme_bw() + guides(color = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("4 groups")
library(RColorBrewer)
mypal = c(brewer.pal(9, "Blues"), brewer.pal(9, "BuGn"),
brewer.pal(9, "Greys"), brewer.pal(9, "PuBu"), brewer.pal(9, "YlGn"))
gp6 = ggplot() +
geom_sf(data = edsm, mapping = aes(fill = Region, color = Region), alpha = 0.5)+
scale_fill_brewer(palette = "Set3")+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
# geom_sf_label(aes(label = Station), stashap) +
scale_color_brewer(palette = "Set1")+
theme_bw() + guides(color = FALSE, fill = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("6 groups")
gp6
gp6a = ggplot() +
geom_sf(data = edsm, mapping = aes(fill = Region), alpha = 0.5)+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
theme_bw() + guides(color = FALSE, fill = FALSE) +
#coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("6 groups")
gp6a
gp12 = ggplot() +
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps12)))+
theme_bw() + guides(color = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("12 groups")
library(gridExtra)
foo = grid.arrange(gp4, gp6, gp12, nrow = 1, ncol = 3)
#ggsave("tempclusters.png",foo, device = "png", width = 12)
```
## Join with EDSM regions
I want to see how closely the EDSM regions map to my clusters. Brian made one version of the shapefile that he tweaked a bit to cover more of the region. THen there is the EDSM subregions, teh EDSM regions, and the EDSM strata. Do any of them match our clusters? I'm using the 6-cluster grouping throughout.I want to see how closely the EDSM regions map to my clusters.
```{r}
spas = st_join(stashap, edsm)
stjoin = as.data.frame(spas)[,1:8]
#read in shapefile of EDSM regions
edsm = read_sf("EDSM_Subregions/EDSM_Subregions_03302020.shp")
edsm2 = read_sf("EDSM_Subregions/Strata_1Dec2018_31Mar2019/Strata_1Dec2018_31Mar2019.shp")
edsm3 = read_sf("EDSM_Subregions/Subregions Phase 1/Subregions_1Dec2018_present.shp")
#reference system is UTMs
st_crs(edsm) = 32610
st_crs(edsm2) = 32610
st_crs(edsm3) = 32610
edsm = st_transform(edsm, 4326)
edsm2 = st_transform(edsm2, 4326)
edsm3 = st_transform(edsm3, 4326)
#perform a spatial join between the stations and Sam's subregions
spas = st_join(stashap, edsm)
spas2 = st_join(stashap, edsm2)
spas3 = st_join(stashap, edsm3)
stjoin = as.data.frame(spas)[,1:8]
stjoin2 = as.data.frame(spas2)[,1:8]
stjoin3 = as.data.frame(spas3)[,1:8]
stjoin = rename(stjoin, SamsSubregion = SubRegion, SamsRegion = Region)
stjoin3 = rename(stjoin3, EDSMSubregion = SubRegion, EDSMRegion = Region)
#put them all together into one master data frame for later use.
allstas = left_join(stjoin, stjoin2) %>%
left_join(stjoin3) %>%
rename(cluster = grps6) %>%
dplyr::select(-geometry, -grps2, -grps4, -grps12, -SQM)
#See how many stations are in each subregion.
#regions = pivot_wider(stjoin, id_cols = c(Region, SubRegion), names_from = grps6, values_from = Station, values_fn = length)
#set up a palette for graphing all the subregions
mypal = c(brewer.pal(9, "Blues"), brewer.pal(9, "BuGn"),
brewer.pal(9, "Greys"), brewer.pal(9, "PuBu"),
brewer.pal(9, "YlGn"), brewer.pal(9, "BuPu"))
#plot of Sam's regions
gp6 = ggplot() +
geom_sf(data = edsm, mapping = aes(fill = Region, color = Region), alpha = 0.5)+
scale_fill_brewer(palette = "Set3")+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
scale_color_brewer(palette = "Set1")+
theme_bw() + guides(color = FALSE, fill = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("Sam's Regions, 6 clusgters")
gp6
#plot of Sam's subregions
gp6x = ggplot() +
geom_sf(data = edsm, mapping = aes(fill = SubRegion), alpha = 0.5)+
scale_fill_manual(values = mypal)+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
theme_bw() + guides(color = FALSE, fill = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("Sam's SubRegions, 6 clusgters")
gp6x
#Plot of EDSM's strata
gp6a = ggplot() +
geom_sf(data = edsm2, mapping = aes(fill = Stratum), alpha = 0.5)+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
theme_bw() + guides(color = FALSE, fill = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("EDSM strata, 6 clusters")
gp6a
#plot of ESM's subregions
gp6b = ggplot() +
geom_sf(data = edsm3, mapping = aes(fill = SubRegion), alpha = 0.5)+
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = as.factor(grps6)))+
theme_bw() + guides(color = FALSE, fill = FALSE) +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("EDSM Subregions, 6 clusters")
gp6b
grid.arrange(gp6, gp6x, gp6a, gp6b, nrow = 1, ncol = 4)
```
OK, Now let's figure out what regions ALL the points are in (even the ones that don't have a complete data set from the past five years..
I used a Leaflet app to draw polygons aroudn the regions.
```{r}
# read in the new region shapefile
RosieReg = read_sf("RosieRegions/shpExport.shp")
#The shiny app didn't actually allow me to add attributes, so I'll do it now
RosieReg$notes = c("Suisun Marsh", "Suisun Bay", "Sac River", "North Delta", "Far North", "San Joaquin", "South")
#perform a spatial join between the stations and my regions
spasR = st_join(stashap, RosieReg)
write.csv(as.data.frame(spasR[,c(1,6,7)]), "Stations_wRosiesReg.csv")
stashap = mutate(stashap, cluster = grps6)
stashap$cluster[which(is.na(stashap$cluster))] = 7
stashap$cluster = factor(stashap$cluster, labels = c("1","2","3","4","5","6","NA"))
mypal2 = c(brewer.pal(6, "Dark2"), "grey")
#plot of my regions
gp6c = ggplot() +
geom_sf(data = RosieReg, mapping = aes(fill = notes), alpha = 0.5)+
# geom_sf_label(aes(label = notes), RosieReg, nudge_y = .1, nudge_x = 0.1) +
geom_sf(data = delta, alpha = 0.5)+
geom_sf(data = stashap, mapping =aes(color = cluster))+
scale_fill_brewer(palette = "Pastel2", guide = NULL) +
scale_color_manual(values = mypal2)+
theme_bw() +
coord_sf(xlim = c(-122.2, -121), ylim = c(37.6, 38.8))+
ggtitle("Rosie's Regions, 6 clusters")
gp6c
regions = pivot_wider(stjoin, id_cols = c(Region, SubRegion), names_from = grps6, values_from = Station, values_fn = length)
```
Can I see how far apart the clusters are?
```{r}
tempmean2 = merge(tempmean, cutday[,c(1,6)])
#put it into wide format for the cluster analysis
tempwide2 = pivot_wider(tempmean2, id_cols = c(Station, grps6),
names_from = julian, values_from = Temp)
row.names(tempwide2) = tempwide2$Station
library(dendextend)
#calculate distance and cluster
tempdist = dist(tempwide2[,c(-1,-2)], "euclidean")
tempfit = hclust(tempdist, method = "ward.D")
plot(tempfit, main = "Clusters based on daily ave temp", cex = 0.6)
d <- as.dendrogram(tempfit)
mypal3 = mypal2[c(4,2,3,1,5,6)]
d <- d %>% color_branches(k=6, col = mypal3) %>% color_labels(k=6, col = mypal3)
plot(d)
```