auxfunc.R 16.3 KB
Newer Older
dmattek's avatar
dmattek committed
1
## Custom plotting
dmattek's avatar
dmattek committed
2
require(ggplot2)
dmattek's avatar
Mod:    
dmattek committed
3
4
5
require(RColorBrewer)
require(gplots) # for heatmap.2
require(grid) # for modifying grob
dmattek's avatar
dmattek committed
6

dmattek's avatar
dmattek committed
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
rhg_cols <- c(
  "#771C19",
  "#AA3929",
  "#E25033",
  "#F27314",
  "#F8A31B",
  "#E2C59F",
  "#B6C5CC",
  "#8E9CA3",
  "#556670",
  "#000000"
)

md_cols <- c(
  "#FFFFFF",
  "#F8A31B",
  "#F27314",
  "#E25033",
  "#AA3929",
  "#FFFFCC",
  "#C2E699",
  "#78C679",
  "#238443"
)

dmattek's avatar
dmattek committed
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
s.cl.linkage = c("ward.D",
                 "ward.D2",
                 "single",
                 "complete",
                 "average",
                 "mcquitty",
                 "centroid")

s.cl.spar.linkage = c("average",
                      "complete", 
                      "single",
                      "centroid")

s.cl.diss = c("euclidean", "maximum", "manhattan", "canberra", "binary", "minkowski")
s.cl.spar.diss = c("squared.distance","absolute.value")

l.col.pal = list(
  "White-Orange-Red" = 'OrRd',
  "Yellow-Orange-Red" = 'YlOrRd',
  "Reds" = "Reds",
  "Oranges" = "Oranges",
  "Greens" = "Greens",
  "Blues" = "Blues",
  "Spectral" = 'Spectral'
)


dmattek's avatar
Mod:    
dmattek committed
59
60
61
62
63
64
65
66
67
68
69
70
myGgplotTraj = function(dt.arg, # data table
                        x.arg,  # string with column name for x-axis
                        y.arg, # string with column name for y-axis
                        group.arg, # string with column name for grouping time series (typicaly cell ID)
                        facet.arg, # string with column name for facetting
                        facet.ncol.arg = 2, # default number of facet columns
                        facet.color.arg = NULL, # vector with list of colours for adding colours to facet names (currently a horizontal line on top of the facet is drawn)
                        line.col.arg = NULL, # string with column name for colouring time series (typically when individual time series are selected in UI)
                        xlab.arg = NULL, # string with x-axis label
                        ylab.arg = NULL, # string with y-axis label
                        plotlab.arg = NULL, # string with plot label
                        dt.stim.arg = NULL, # plotting additional dataset; typically to indicate stimulations (not fully implemented yet, not tested!)
dmattek's avatar
dmattek committed
71
                        tfreq.arg = 1,
dmattek's avatar
dmattek committed
72
                        ylim.arg = NULL,
dmattek's avatar
dmattek committed
73
                        stim.bar.height.arg = 0.1,
dmattek's avatar
Added:    
dmattek committed
74
                        stim.bar.width.arg = 0.5,
dmattek's avatar
Mod:    
dmattek committed
75
                        aux.label1 = NULL, # 1st point label; used for interactive plotting; displayed in the tooltip; typically used to display values of column holding x & y coordinates
dmattek's avatar
Added:    
dmattek committed
76
77
78
79
80
81
                        aux.label2 = NULL,
                        stat.arg = c('', 'mean', 'CI', 'SE')) {
  
  # match arguments for stat plotting
  loc.stat = match.arg(stat.arg, several.ok = TRUE)

dmattek's avatar
Added:    
dmattek committed
82
83
  
  # aux.label12 are required for plotting XY positions in the tooltip of the interactive (plotly) graph
dmattek's avatar
dmattek committed
84
85
  p.tmp = ggplot(dt.arg,
                 aes_string(x = x.arg,
dmattek's avatar
dmattek committed
86
                            y = y.arg,
dmattek's avatar
Added:    
dmattek committed
87
88
89
                            group = group.arg,
                            label  = aux.label1,
                            label2 = aux.label2))
dmattek's avatar
dmattek committed
90
  
dmattek's avatar
dmattek committed
91
92
93
94
95
96
97
98
99
100
101
102
103
  if (is.null(line.col.arg)) {
    p.tmp = p.tmp +
      geom_line(alpha = 0.25, 
                              size = 0.25)
  }
  else {
    p.tmp = p.tmp + 
      geom_line(aes_string(colour = line.col.arg), 
                              alpha = 0.5, 
                              size = 0.5) +
      scale_color_manual(name = '', 
                         values =c("FALSE" = rhg_cols[7], "TRUE" = rhg_cols[3], "SELECTED" = 'green', "NOT SEL" = rhg_cols[7]))
  }
dmattek's avatar
Mod:    
dmattek committed
104
105
106
107
108
109
110
111
112
113

  # this is temporary solution for adding colour according to cluster number
  # use only when plotting traj from clustering!
  # a horizontal line is added at the top of data
  if (!is.null(facet.color.arg)) {

    loc.y.max = max(dt.arg[, c(y.arg), with = FALSE])
    loc.dt.cl = data.table(xx = 1:length(facet.color.arg), yy = loc.y.max)
    setnames(loc.dt.cl, 'xx', facet.arg)
    
dmattek's avatar
Fixed:    
dmattek committed
114
115
    # adjust facet.color.arg to plot
    
dmattek's avatar
Mod:    
dmattek committed
116
117
118
119
120
    p.tmp = p.tmp +
      geom_hline(data = loc.dt.cl, colour = facet.color.arg, yintercept = loc.y.max, size = 4) +
      scale_colour_manual(values = facet.color.arg,
                          name = '')
  }
dmattek's avatar
dmattek committed
121
  
dmattek's avatar
Added:    
dmattek committed
122
123
  if ('mean' %in% loc.stat)
    p.tmp = p.tmp + 
dmattek's avatar
dmattek committed
124
125
126
    stat_summary(
      aes_string(y = y.arg, group = 1),
      fun.y = mean,
dmattek's avatar
Added:    
dmattek committed
127
      colour = 'red',
dmattek's avatar
dmattek committed
128
129
130
131
      linetype = 'solid',
      size = 1,
      geom = "line",
      group = 1
dmattek's avatar
Added:    
dmattek committed
132
133
134
135
136
137
138
139
    )

  if ('CI' %in% loc.stat)
    p.tmp = p.tmp + 
    stat_summary(
      aes_string(y = y.arg, group = 1),
      fun.data = mean_cl_normal,
      colour = 'red',
dmattek's avatar
Mod:    
dmattek committed
140
      alpha = 0.25,
dmattek's avatar
Added:    
dmattek committed
141
142
143
144
145
146
147
148
149
150
      geom = "ribbon",
      group = 1
    )
  
  if ('SE' %in% loc.stat)
    p.tmp = p.tmp + 
    stat_summary(
      aes_string(y = y.arg, group = 1),
      fun.data = mean_se,
      colour = 'red',
dmattek's avatar
Mod:    
dmattek committed
151
      alpha = 0.25,
dmattek's avatar
Added:    
dmattek committed
152
153
154
155
156
157
158
      geom = "ribbon",
      group = 1
    )
  
  
  
  p.tmp = p.tmp + 
dmattek's avatar
dmattek committed
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
    facet_wrap(as.formula(paste("~", facet.arg)),
               ncol = facet.ncol.arg,
               scales = "free_x")
  
  if(!is.null(dt.stim.arg)) {
    p.tmp = p.tmp + geom_segment(data = dt.stim.arg,
                                 aes(x = Stimulation_time - tfreq.arg,
                                     xend = Stimulation_time - tfreq.arg,
                                     y = ylim.arg[1],
                                     yend = ylim.arg[1] + abs(ylim.arg[2] - ylim.arg[1]) * stim.bar.height.arg),
                                 colour = rhg_cols[[3]],
                                 size = stim.bar.width.arg,
                                 group = 1) 
  }
  
dmattek's avatar
dmattek committed
174
175
176
  if (!is.null(ylim.arg)) 
    p.tmp = p.tmp + coord_cartesian(ylim = ylim.arg)
  
dmattek's avatar
dmattek committed
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
  p.tmp = p.tmp + 
    xlab(paste0(xlab.arg, "\n")) +
    ylab(paste0("\n", ylab.arg)) +
    ggtitle(plotlab.arg) +
    theme_bw(base_size = 18, base_family = "Helvetica") +
    theme(
      panel.grid.minor = element_blank(),
      panel.grid.major = element_blank(),
      panel.border = element_blank(),
      axis.line.x = element_line(color = "black", size = 0.25),
      axis.line.y = element_line(color = "black", size = 0.25),
      axis.text.x = element_text(size = 12),
      axis.text.y = element_text(size = 12),
      strip.text.x = element_text(size = 14, face = "bold"),
      strip.text.y = element_text(size = 14, face = "bold"),
      strip.background = element_blank(),
      legend.key = element_blank(),
      legend.key.height = unit(1, "lines"),
      legend.key.width = unit(2, "lines"),
      legend.position = "top"
    )
  
dmattek's avatar
Mod:    
dmattek committed
199
200
201
202
  

  
  return(p.tmp)
dmattek's avatar
dmattek committed
203
204
205
206
207
208
}


userDataGen <- function() {  
  cat(file=stderr(), 'userDataGen: in\n')
  
dmattek's avatar
dmattek committed
209
  locNtp = 40
210
  locNtracks = 100
dmattek's avatar
dmattek committed
211
  locNsites = 4
212
213
  locNwells = 1
  
dmattek's avatar
Mod:    
dmattek committed
214
215
216
217
  x.rand.1 = c(rnorm(locNtp * locNtracks * locNsites * 0.5, .5, 0.1), rnorm(locNtp * locNtracks * locNsites * 0.5, 1, 0.2))
  x.rand.2 = c(rnorm(locNtp * locNtracks * locNsites * 0.5, 0.25, 0.1), rnorm(locNtp * locNtracks * locNsites * 0.5, 0.5, 0.2))
#  x.rand.3 = rep(rnorm(locNtracks, 2, 0.5), 1, each = locNtp)
#  x.rand.4 = rep(rnorm(locNtracks, 1, 0.1), 1, each = locNtp)
218
  
dmattek's avatar
Mod:    
dmattek committed
219
220
#  x.arg = rep(seq(0, locNtp-1) / locNtp * 4 * pi, locNtracks * locNsites)
  x.arg = rep(seq(1, locNtp), locNtracks * locNsites)
dmattek's avatar
dmattek committed
221
222
223
  
  dt.nuc = data.table(Metadata_Site = rep(1:locNsites, each = locNtp * locNtracks),
                      Metadata_Well = rep(1:locNwells, each = locNtp * locNsites * locNtracks / locNwells),
224
                      Metadata_RealTime = x.arg,
dmattek's avatar
Mod:    
dmattek committed
225
226
227
228
229
                      objCyto_Intensity_MeanIntensity_imErkCor = x.rand.1,
                      objNuc_Intensity_MeanIntensity_imErkCor  = x.rand.2,
                      objNuc_Location_X = runif(locNtp * locNtracks * locNsites, min = 0, max = 1),
                      objNuc_Location_Y = runif(locNtp * locNtracks * locNsites, min = 0, max = 1),
#                      objCyto_Intensity_MeanIntensity_imErkCor = x.rand.3 + ifelse(x.arg < 4, 0, 1) / x.rand.3,
230
#                      objNuc_Intensity_MeanIntensity_imErkCor  = c(rnorm(locNtp * locNtracks * locNsites * 0.5, .25, 0.1), rnorm(locNtp * locNtracks * locNsites * 0.5, .5, 0.2)),
dmattek's avatar
dmattek committed
231
232
233
234
                      TrackLabel = rep(1:(locNtracks*locNsites), each = locNtp))
  
  return(dt.nuc)
}
dmattek's avatar
dmattek committed
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


# Returns original dt with an additional column with normalized quantity.
# The column to be normalised is given by 'in.meas.col'.
# The name of additional column is the same as in.meas.col but with ".norm" suffix added.
# Normalisation is based on part of the trajectory;
# this is defined by in.rt.min and max, and the column with time in.rt.col.
# Additional parameters:
# in.by.cols - character vector with 'by' columns to calculate normalisation per group
#              if NULL, no grouping is done
# in.robust - whether robust measures should be used (median instead of mean, mad instead of sd)
# in.type - type of normalization: z.score or mean (fi.e. old change w.r.t. mean)

myNorm = function(in.dt,
                  in.meas.col,
                  in.rt.col = 'RealTime',
                  in.rt.min = 10,
                  in.rt.max = 20,
                  in.by.cols = NULL,
                  in.robust = TRUE,
                  in.type = 'z.score') {
  loc.dt <-
    copy(in.dt) # copy so as not to alter original dt object w intermediate assignments
  
  if (is.null(in.by.cols)) {
    if (in.robust)
dmattek's avatar
Fixed:    
dmattek committed
261
262
      loc.dt.pre.aggr = loc.dt[get(in.rt.col) >= in.rt.min &
                                 get(in.rt.col) <= in.rt.max, .(meas.md = median(get(in.meas.col), na.rm = TRUE),
dmattek's avatar
dmattek committed
263
264
                                                               meas.mad = mad(get(in.meas.col), na.rm = TRUE))]
    else
dmattek's avatar
Fixed:    
dmattek committed
265
266
      loc.dt.pre.aggr = loc.dt[get(in.rt.col) >= in.rt.min &
                                 get(in.rt.col) <= in.rt.max, .(meas.md = mean(get(in.meas.col), na.rm = TRUE),
dmattek's avatar
dmattek committed
267
268
269
270
271
                                                               meas.mad = sd(get(in.meas.col), na.rm = TRUE))]
    
    loc.dt = cbind(loc.dt, loc.dt.pre.aggr)
  }  else {
    if (in.robust)
dmattek's avatar
Fixed:    
dmattek committed
272
273
      loc.dt.pre.aggr = loc.dt[get(in.rt.col) >= in.rt.min &
                                 get(in.rt.col) <= in.rt.max, .(meas.md = median(get(in.meas.col), na.rm = TRUE),
dmattek's avatar
dmattek committed
274
275
                                                               meas.mad = mad(get(in.meas.col), na.rm = TRUE)), by = in.by.cols]
    else
dmattek's avatar
Fixed:    
dmattek committed
276
277
      loc.dt.pre.aggr = loc.dt[get(in.rt.col) >= in.rt.min &
                                 get(in.rt.col) <= in.rt.max, .(meas.md = mean(get(in.meas.col), na.rm = TRUE),
dmattek's avatar
dmattek committed
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
                                                               meas.mad = sd(get(in.meas.col), na.rm = TRUE)), by = in.by.cols]
    
    loc.dt = merge(loc.dt, loc.dt.pre.aggr, by = in.by.cols)
  }
  
  
  if (in.type == 'z.score') {
    loc.dt[, meas.norm := (get(in.meas.col) - meas.md) / meas.mad]
  } else {
    loc.dt[, meas.norm := (get(in.meas.col) / meas.md)]
  }
  
  setnames(loc.dt, 'meas.norm', paste0(in.meas.col, '.norm'))
  
  loc.dt[, c('meas.md', 'meas.mad') := NULL]
  return(loc.dt)
dmattek's avatar
dmattek committed
294
295
}

dmattek's avatar
dmattek committed
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
385
386
387
388
389
390
391
392
393
394
# Plots a scatter plot with marginal histograms
# Points are connected by a line (grouping by cellID)
#
# Assumes an input of data.table with
# x, y - columns with x and y coordinates
# id - a unique point identifier (here corresponds to cellID)
# mid - a (0,1) column by which points are coloured (here corresponds to whether cells are within bounds)

myGgplotScat = function(dt.arg,
                        band.arg = NULL,
                        facet.arg = NULL,
                        facet.ncol.arg = 2,
                        xlab.arg = NULL,
                        ylab.arg = NULL,
                        plotlab.arg = NULL,
                        alpha.arg = 1,
                        group.col.arg = NULL) {
  p.tmp = ggplot(dt.arg, aes(x = x, y = y))
  
  if (is.null(group.col.arg)) {
    p.tmp = p.tmp +
      geom_point(alpha = alpha.arg, aes(group = id))
  } else {
    p.tmp = p.tmp +
      geom_point(aes(colour = as.factor(get(group.col.arg)), group = id), alpha = alpha.arg) +
      geom_path(aes(colour = as.factor(get(group.col.arg)), group = id), alpha = alpha.arg) +
      scale_color_manual(name = group.col.arg, values =c("FALSE" = rhg_cols[7], "TRUE" = rhg_cols[3], "SELECTED" = 'green'))
  }
  
  if (is.null(band.arg))
    p.tmp = p.tmp +
      stat_smooth(
        method = function(formula, data, weights = weight)
          rlm(formula, data, weights = weight, method = 'MM'),
        fullrange = FALSE,
        level = 0.95,
        colour = 'blue'
      )
  else {
    p.tmp = p.tmp +
      geom_abline(slope = band.arg$a, intercept = band.arg$b) +
      geom_abline(
        slope = band.arg$a,
        intercept =  band.arg$b + abs(band.arg$b)*band.arg$width,
        linetype = 'dashed'
      ) +
      geom_abline(
        slope = band.arg$a,
        intercept = band.arg$b - abs(band.arg$b)*band.arg$width,
        linetype = 'dashed'
      )
  }
  
  if (!is.null(facet.arg)) {
    p.tmp = p.tmp +
      facet_wrap(as.formula(paste("~", facet.arg)),
                 ncol = facet.ncol.arg)
    
  }
  
  
  if (!is.null(xlab.arg))
    p.tmp = p.tmp +
      xlab(paste0(xlab.arg, "\n"))
  
  if (!is.null(ylab.arg))
    p.tmp = p.tmp +
      ylab(paste0("\n", ylab.arg))
  
  if (!is.null(plotlab.arg))
    p.tmp = p.tmp +
      ggtitle(paste0(plotlab.arg, "\n"))
  
  
  
  p.tmp = p.tmp +
    theme_bw(base_size = 18, base_family = "Helvetica") +
    theme(
      panel.grid.minor = element_blank(),
      panel.grid.major = element_blank(),
      axis.line.x = element_line(color = "black", size = 0.25),
      axis.line.y = element_line(color = "black", size = 0.25),
      axis.text.x = element_text(size = 12),
      axis.text.y = element_text(size = 12),
      strip.text.x = element_text(size = 14, face = "bold"),
      strip.text.y = element_text(size = 14, face = "bold"),
      strip.background = element_blank(),
      legend.key = element_blank(),
      legend.key.height = unit(1, "lines"),
      legend.key.width = unit(2, "lines"),
      legend.position = "none"
    )
  
  # Marginal distributions don;t work with plotly...
  # if (is.null(facet.arg))
  #   ggExtra::ggMarginal(p.scat, type = "histogram",  bins = 100)
  # else
  return(p.tmp)
}
dmattek's avatar
dmattek committed
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410

myGgplotTheme = theme_bw(base_size = 18, base_family = "Helvetica") +
  theme(
    panel.grid.minor = element_blank(),
    panel.grid.major = element_blank(),
    axis.line.x = element_line(color = "black", size = 0.25),
    axis.line.y = element_line(color = "black", size = 0.25),
    axis.text.x = element_text(size = 12, angle = 45, hjust = 1),
    axis.text.y = element_text(size = 12),
    strip.text.x = element_text(size = 14, face = "bold"),
    strip.text.y = element_text(size = 14, face = "bold"),
    strip.background = element_blank(),
    legend.key = element_blank(),
    legend.key.height = unit(1, "lines"),
    legend.key.width = unit(2, "lines"),
    legend.position = "right"
dmattek's avatar
Mod:    
dmattek committed
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
  )


myPlotHeatmap <- function(data.arg,
                          dend.arg,
                          palette.arg,
                          palette.rev.arg = TRUE,
                          dend.show.arg = TRUE,
                          key.show.arg = TRUE,
                          margin.x.arg = 5,
                          margin.y.arg = 20,
                          nacol.arg = 0.5,
                          colCol.arg = NULL,
                          labCol.arg = NULL,
                          font.row.arg = 1,
                          font.col.arg = 1,
                          title.arg = 'Clustering') {
  
  if (palette.rev.arg)
    my_palette <-
    rev(colorRampPalette(brewer.pal(9, palette.arg))(n = 99))
  else
    my_palette <-
    colorRampPalette(brewer.pal(9, palette.arg))(n = 99)
  
  
  col_labels <- get_leaves_branches_col(dend.arg)
  col_labels <- col_labels[order(order.dendrogram(dend.arg))]
  
  if (dend.show.arg) {
    assign("var.tmp.1", dend.arg)
    var.tmp.2 = "row"
  } else {
    assign("var.tmp.1", FALSE)
    var.tmp.2 = "none"
  }
  
  loc.p = heatmap.2(
    data.arg,
    Colv = "NA",
    Rowv = var.tmp.1,
    srtCol = 90,
    dendrogram = var.tmp.2,
    trace = "none",
    key = key.show.arg,
    margins = c(margin.x.arg, margin.y.arg),
    col = my_palette,
    na.col = grey(nacol.arg),
    denscol = "black",
    density.info = "density",
    RowSideColors = col_labels,
    colRow = col_labels,
    colCol = colCol.arg,
    labCol = labCol.arg,
    #      sepcolor = grey(input$inPlotHierGridColor),
    #      colsep = 1:ncol(loc.dm),
    #      rowsep = 1:nrow(loc.dm),
    cexRow = font.row.arg,
    cexCol = font.col.arg,
    main = title.arg
  )
  
  return(loc.p)
}