This repository was archived by the owner on Feb 1, 2025. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathsummarize_table_function.R
executable file
·528 lines (449 loc) · 29.1 KB
/
summarize_table_function.R
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
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
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
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
################################
# Name: MacIntosh Cornwell
# Email: mcornwell1957@gmail.com
################################
## Function to tabulate and summarize an input data - will return counts and also the stats testing the significance of the diff between groups
# Get the functions from github
library(devtools)
source_url('https://raw.githubusercontent.com/mattmuller0/Rtools/main/mgc_plotting_functions.R')
## EXAMPLE
# summarystatfile_seq <- "/Users/tosh/Desktop/Ruggles_Lab/projects/ischemia2020/data/summarystats_IMGDEGIS_20200123.csv"
# sumstattab_seq <- summarize_table(intable = meta1[meta1[,"sequencing"] %in% "yes",], groupvar = "IMGDEGIS",
# outfile = summarystatfile_seq, calc_stats = FALSE)
## A better summary function that will summarize table stats - first column should in general be some kind of ID
# Input: intable - the intable (first col is probably some kind of ID)
# summarystatfile - the file to write out too, this is a messy output, but is pretty good considering..
# will write out the file, and also output the summary list with each column as an entry in thelist
#### This is horrible coding, have to clean up at some point, but is working as intended now, so oh well.
summarize_table <- function(intable, groupvar = NULL, outfile, calc_stats = FALSE, calc_ci = FALSE) {
## Assign our infotab as the intable either as is, or with the grouping column excluded
if (!is.null(groupvar)){
groupcol <- intable[,groupvar,drop=FALSE]
infotab <- intable[,!colnames(intable) %in% groupvar]
subgroupvars <- sort(as.character(unique(groupcol[,groupvar])))
} else {
subgroupvars <- "all"
infotab <- intable
}
## For each column, we have to run analysis
summarystatlist <- list()
for (columnnum in seq_len(ncol(infotab))){
## For each subgroup, we have to tabulate our info
## If there is no goruping column, then we just run through once
outstatlist <- list()
for (groupvarnum in seq_len(length(subgroupvars))) {
## Grab the col
if (!is.null(groupvar)){
subtab <- infotab[groupcol[,groupvar] %in% subgroupvars[groupvarnum],columnnum,drop = FALSE]
} else {
subtab <- infotab[,columnnum,drop=FALSE]
}
statlabel <- colnames(subtab)
## This is where we actually do the tabulation
## if first column - thats the ID column, so just return the total and the unique
## if not, then see if column is character - then count, if number - then return statistical summary
if (columnnum == 1){
outstatlist[[groupvarnum]] <- cbind(classes = c("total", "unique_entries"), values = c(nrow(subtab), length(unique(subtab[,1]))))
} else {
if (class(subtab[,1]) == "character" | class(subtab[,1]) == "factor" | class(subtab[,1]) == "logical"){
## failsafe for when the second group is all NAs
if (sum(is.na(subtab[,1])) == length(subtab[,1])) {
outstatlist[[groupvarnum]] <- data.frame(classes = "NAvalue", values = sum(is.na(subtab[,1])))
} else {
outstatlist[[groupvarnum]] <- data.frame(table(subtab[,1]))
colnames(outstatlist[[groupvarnum]]) <- c("classes", "values")
}
}
if (class(subtab[,1]) == "numeric" | class(subtab[,1]) == "integer") {
tempsum <- c(summary(subtab[,1]), sd = sd(subtab[,1], na.rm = TRUE))
outstatlist[[groupvarnum]] <- data.frame(ids = names(tempsum), values = as.numeric(unname(tempsum)))
colnames(outstatlist[[groupvarnum]]) <- c("classes", "values")
}
}
names(outstatlist)[groupvarnum] <- subgroupvars[groupvarnum]
}
## Merge our results back together
outstat <- suppressWarnings(Reduce(function(dtf1, dtf2) merge(dtf1, dtf2, by = "classes", all = TRUE, sort = FALSE), c(outstatlist)))
outstat[is.na(outstat)] <- 0
## Changing the naming convention if we are grouping by a category
if (is.null(groupvar)){
colnames(outstat) <- c("classes", subgroupvars)
} else {
colnames(outstat) <- c(paste0(groupvar, "_cat"), subgroupvars)
}
## Calculate the stats on our comparison (if applicable)
if (calc_stats == TRUE) {
if (columnnum == 1) {
outstat <- outstat
} else {
if (class(infotab[,columnnum]) == "character" | class(infotab[,columnnum]) == "factor") {
if (length(subgroupvars) == 2) {
stattab <- outstat[,c(2:ncol(outstat))]
rownames(stattab) <- outstat[,1]
statval <- apply(stattab, 1, function(x) {
group1 <- c(rep(1,x[1]), rep(0, unname(colSums(stattab))[1] - x[1]))
group2 <- c(rep(1,x[2]), rep(0, unname(colSums(stattab))[2] - x[2]))
if(mean(group1) == mean(group2)) {
1
} else {
## failsafe for case where you have 2 separate pops compared (all 1s vs all 0s)
## Also need failsafe for when you have a group of size 1
if (length(unique(group1)) == 1 & length(unique(group2)) == 1 |
length(group1) == 1 | length(group2) == 1) {
pvalout <- 0
if (calc_ci == TRUE) {
ci_val <- "0 (0 - 0)"
c(pvalout, ci_val)
} else {
pvalout
}
} else {
pvalout <- t.test(group1, group2)$p.value
## 20220228 - adding in a CI?
if (calc_ci == TRUE) {
grab_ci <- tryCatch(t.test(group1, group2)$conf.int,
error = function(e) c(0, 0))
diff_val <- t.test(group1, group2)$estimate[1] / t.test(group1, group2)$estimate[2]
ci_val <- paste0(diff_val, " (", paste(grab_ci, collapse = " - "), ")")
c(pvalout, ci_val)
} else {
pvalout
}
}
}
})
# outstat[,"statval"] <- statval
if (calc_ci == TRUE) {
outstat[,c("statval", "ci_val")] <- t(statval)
} else{
outstat[,"statval"] <- statval
}
outstat <- outstat[order(factor(outstat[,1], levels = outstat[,1][order(as.character(outstat[,1]))])),]
} else {
## NOT WRITTEN YET, WILL NEED TO WRITE AS APPLICABLE
# NEED TO DO A CHISQUARED TEST HERE!!!!
# aovtab <- intable[groupcol[,groupvar] %in% subgroupvars,c(statlabel, groupvar),drop = FALSE]
# aovtab[sapply(aovtab, is.character)] <- lapply(aovtab[sapply(aovtab, is.character)], as.factor)
# aovtab[sapply(aovtab, is.factor)] <- lapply(aovtab[sapply(aovtab, is.factor)],
# function(x) `levels<-`(addNA(x), c(levels(x), "NA")))
# statval <- kruskal.test(as.formula(paste0(groupvar, "~", ".")), aovtab)$p.value
# statval <- chisq.test(groupcol[,1], infotab[,columnnum], simulate.p.value = TRUE)$p.value
statval <- tryCatch(chisq.test(groupcol[,1], infotab[,columnnum], simulate.p.value = TRUE)$p.value,
error = function(e) NA)
outstat[,"statval"] <- statval
outstat <- outstat[order(factor(outstat[,1], levels = outstat[,1][order(as.character(outstat[,1]))])),]
}
}
if (class(infotab[,columnnum]) == "numeric" | class(infotab[,columnnum]) == "integer") {
## if two, then t test
if (length(subgroupvars) == 2) {
group1 <- infotab[groupcol[,groupvar] %in% subgroupvars[1],columnnum,drop = FALSE]
group2 <- infotab[groupcol[,groupvar] %in% subgroupvars[2],columnnum,drop = FALSE]
if (length(group1[!is.na(group1)]) < 3 | length(group2[!is.na(group2)]) < 3) {
statval <- "toofew"
ci_val <- "toofew"
} else {
## failsafe for case where you have 2 separate pops compared (all 1s vs all 0s)
if (nrow(unique(group1)) == 1 & nrow(unique(group2)) == 1) {
statval <- 0
} else {
## Need a tryCatch here too for when you have data that is all 1s and NAs vs eachother (or other equiv scenario
statval <- tryCatch(t.test(group1, group2)$p.value,
error = function(e) NA)
# statval <- t.test(group1, group2)$p.value
## 20220228 - adding in a CI?
grab_ci <- tryCatch(t.test(group1, group2)$conf.int,
error = function(e) c(0, 0))
diff_val <- t.test(group1, group2)$estimate[1] / t.test(group1, group2)$estimate[2]
ci_val <- paste0(diff_val, " (", paste(grab_ci, collapse = " - "), ")")
}
}
outstat[,"statval"] <- c(statval, rep("", nrow(outstat)-1))
if (calc_ci == TRUE) {outstat[,"ci_val"] <- c(ci_val, rep("", nrow(outstat)-1))}
} else {
## NOT WRITTEN YET, WILL NEED TO WRITE AS APPLICABLE
# # if three or more, melt down into groups, and then anova test (melt so 2 cols, 1 with group 1 with stat)
# res.aov <- aov(var ~ group, data = melted data
# summary(res.aov)
aovtab <- intable[groupcol[,groupvar] %in% subgroupvars,c(statlabel, groupvar),drop = FALSE]
# an anova is finicky and needs to have the levels be real values (which isnt surprising) - so issues will happen with blanks and NAs, and there may be times where there are blanks AND NAs, so for this case, lets change them
aovtab[,groupvar][aovtab[,groupvar] == ""] <- "blankval"
aovtab[,groupvar][is.na(aovtab[,groupvar])] <- "NAval"
aovtab <- na.omit(aovtab)
## Failsafe to get rid of NA values and turn into string which can then be factors
#`levels<-`(addNA(aovtab[,1]), c(levels(aovtab[,1]), "NA"))
aovtab[sapply(aovtab, is.character)] <- lapply(aovtab[sapply(aovtab, is.character)], as.factor)
# aovtab[sapply(aovtab, is.factor)] <- lapply(aovtab[sapply(aovtab, is.factor)],
# function(x) `levels<-`(addNA(x), c(levels(x), "NA")))
if (nrow(aovtab) > 2) { ## Failsafe added 2021-9-14
# aov formula doesnt like dashes! So we have to change that: 2022-05-02
statlabel <- gsub("-", "_", statlabel)
colnames(aovtab)[1] <- gsub("-", "_", statlabel)
aovform <- as.formula(paste0(statlabel, "~", "."))
aovout <- aov(aovform, aovtab)
statval <- summary(aovout)[[1]][["Pr(>F)"]][[1]]
outstat[,"statval"] <- statval
outstat <- outstat[order(factor(outstat[,1], levels = outstat[,1][order(as.character(outstat[,1]))])),]
} else {
outstat[,"statval"] <- NA
}
}
}
}
}
# outstat <- outstat[order(factor(outstat[,1], levels = outstat[,1][order(as.character(outstat[,1]))])),]
## Added a sorting function in here to help clean things up
## Write out the statlist and append to the table
summarystatlist[[columnnum]] <- outstat
names(summarystatlist)[columnnum] <- statlabel
## Adding a blank line to make the table a little easier to read?
outstat <- rbind(outstat, rep(NA,nrow(outstat)))
write.table(statlabel, outfile, sep = ",", append = columnnum != 1, row.names = FALSE, col.names = FALSE, quote = FALSE, na = "")
suppressWarnings(write.table(outstat, outfile, sep = ",", append = TRUE, row.names = FALSE, col.names = !is.null(groupvar), quote = FALSE, na = ""))
}
return(summarystatlist)
}
## Play coding - turning my summary table output into a clean table using kable and a lot of formatting
## Ok - so we need to think about things that we want to do to make it easier
# 1 - move categories into the first column
# 2 - add percentages to the values
# 3 - round the stat values
# 4 - combining all into one big table
## NOT DOING the following
# picking which features to keep (just list all, should be easy to edit later)
#
# sumstattable_input = sumstattab_seq
# addpercents = "vertical" # adds percents by category (as opposed to by feature)
# contsummary = c("mean", "sd") # what stats to pick for the continuous variable summary
# roundpvaldigits = 3
# c("median", "iqr")
clean_summarize_table_output <- function(sumstattable_input,
addpercents = "vertical", contsummary = c("mean", "sd"), roundpvaldigits = 3) {
## Run over every entry in the list, and then clean as applicable:
outtable_list <- list()
for (stattablecatnum in seq_len(length(sumstattable_input))) {
stattablecat_label <- names(sumstattable_input)[stattablecatnum]
stattable_sel <- sumstattable_input[[stattablecatnum]]
subgroups <- colnames(stattable_sel)[2:ncol(stattable_sel)][
!grepl("statval|ci_val", colnames(stattable_sel)[2:ncol(stattable_sel)])]
## Ok, so the first one SHOULD be the ID number, and if it is, grab these values for later and move on
if (identical(stattable_sel[,1], c("total", "unique_entries"))) {
cohorttotals <- stattable_sel[stattable_sel[,1] == "total",2:ncol(stattable_sel)]
cohorttotals_replacementtable <- data.frame(values = t(cohorttotals),
original = names(cohorttotals),
replacement = paste0(names(cohorttotals), " (N = ", cohorttotals, ")"))
next
}
## Next - if we have a continuous variable, then we will have the same suumary stats, so we can detect by that, and then adjust accordingly
if (sum(stattable_sel[,1] %in% c("Min.", "1st Qu.", "Median", "Mean", "3rd Qu.", "Max.", "sd")) >= 7) { ## we have a cont variable
# if(contsummary[1] == "mean") {var1 <- stattable_sel[stattable_sel[,1] == "Mean",2:(ncol(stattable_sel)-1)]}
# if(contsummary[1] == "median") {var1 <- stattable_sel[stattable_sel[,1] == "Median",2:(ncol(stattable_sel)-1)]}
# if(contsummary[2] == "sd") {var2 <- stattable_sel[stattable_sel[,1] == "sd",2:(ncol(stattable_sel)-1)]}
# if(contsummary[2] == "iqr") {var2 <- stattable_sel[stattable_sel[,1] == "1st Qu.",2:(ncol(stattable_sel)-1)]}
if(contsummary[1] == "mean") {var1 <- stattable_sel[stattable_sel[,1] == "Mean",subgroups]}
if(contsummary[1] == "median") {var1 <- stattable_sel[stattable_sel[,1] == "Median",subgroups]}
if(contsummary[2] == "sd") {var2 <- round(stattable_sel[stattable_sel[,1] == "sd",subgroups],2)}
if(contsummary[2] == "iqr") {var2 <- paste0(
round(stattable_sel[stattable_sel[,1] == "1st Qu.",subgroups], 2), "-",
round(stattable_sel[stattable_sel[,1] == "3rd Qu.",subgroups],2))
}
outvar <- paste0(round(var1,2), " (", var2, ")")
names(outvar) <- names(var1)
outstat <- suppressWarnings(round(as.numeric(stattable_sel[1,"statval"]), roundpvaldigits))
if (length(outstat) > 0){
if (outstat %in% 0) {outstat <- "<0.001"}
} else{
outstat <- NULL
}
outtable <- cbind(
# category = paste0(stattablecat_label, ", ", contsummary[1], " (±", contsummary[2], ")"),
category = stattablecat_label,
feature = paste0(contsummary[1], " (±", contsummary[2], ")"),
t(data.frame(outvar)),
statval = outstat)
if (ncol(outtable) == 3) {colnames(outtable) <- c("category", "feature", "all")}
} else { ## THIS SHOULD BE CATEGORICAL, I CANT THINK OF ANYTHING ELSE THIS COULD BE RIGHT NOW...
outtable <- cbind(category = stattablecat_label, stattable_sel)
## If we have a multigroup summary with stats do this:
if (sum(grepl("statval", colnames(stattable_sel))) > 0) {
colnames(outtable) <- c("category", "feature", colnames(stattable_sel)[2:(ncol(stattable_sel)-1)], "statval")
outtable[,"statval"] <- round(as.numeric(outtable[,"statval"]), roundpvaldigits)
outtable[,"statval"] <- ifelse(outtable[,"statval"] == 0, "<0.001", outtable[,"statval"])
} else {
## Lastly - if we have a single group summary and no stats, then do this:
if (ncol(stattable_sel) == 2){
outtable <- cbind(category = stattablecat_label, stattable_sel)
colnames(outtable) <- c("category", "feature", "all")
} else { ## If we have multigroup with NO stats do this:
outtable <- cbind(category = stattablecat_label, stattable_sel)
colnames(outtable) <- c("category", "feature", colnames(stattable_sel)[2:ncol(stattable_sel)])
}
}
## Now add some things
if (!is.null(addpercents)) {
outtable[,"feature"] <- paste0(outtable[,"feature"], ", N (%)")
## percents by row or column?
percentorderparam <- ifelse(addpercents == "vertical", 2, 1)
# outtable[,colnames(stattable_sel)[2:(ncol(stattable_sel)-1)]] <- apply(
# outtable[,colnames(stattable_sel)[2:(ncol(stattable_sel)-1)]], percentorderparam, function(x) {
# percentvals <- round(x/sum(x),3)*100
# outvals <- paste0(x, " (", percentvals, ")")
# outvals
# }) ## TOSH CHANGED - the minus one correction needs to be changed, just want to grab all groups - all for no groups
tempout <- apply(outtable[,subgroups,drop=FALSE], percentorderparam, function(x) {
percentvals <- round(x/sum(x),3)*100
outvals <- paste0(x, " (", percentvals, ")")
outvals
})
if (percentorderparam == 1) {tempout <- t(tempout)}
outtable[,subgroups] <- tempout
} else{
outtable[,"feature"] <- paste0(outtable[,"feature"], ", N")
}
}
outtable_list[[stattablecatnum]] <- outtable
}
combotable <- do.call(rbind, outtable_list)
rownames(combotable) <- NULL
colnames(combotable)[3:(ncol(sumstattable_input[[1]])+1)] <- cohorttotals_replacementtable$replacement[match(colnames(combotable)[3:(ncol(sumstattable_input[[1]])+1)], cohorttotals_replacementtable$original)]
return(combotable)
}
## A better summary function that will summarize table stats - first column should in general be some kind of ID
# Input: intable - the intable (first col is probably some kind of ID)
# summarystatfile - the file to write out too, this is a messy output, but is pretty good considering..
# will write out the file, and also output the summary list with each column as an entry in thelist
#### This is horrible coding, have to clean up at some point, but is working as intended now, so oh well.
summarize_and_correlate_table <- function(intable, groupvar) {
## Assign our infotab as the intable either as is, or with the grouping column excluded
groupcol <- intable[,groupvar,drop=FALSE]
infotab <- intable[,!colnames(intable) %in% groupvar]
## For each column, we have to run analysis
outstatlist <- list()
for (columnnum in seq_len(ncol(infotab))){
# for (columnnum in seq_len(11)){
subtab <- infotab[,columnnum,drop=FALSE]
statlabel <- colnames(subtab)
## If first column - then assume IDs, and just count them up
if (columnnum == 1){
outstatlist[[columnnum]] <- data.frame(rval = nrow(subtab), pval = length(unique(subtab[,1])), row.names = statlabel)
} else {
if (class(subtab[,1]) == "character" | class(subtab[,1]) == "factor"){
## NOT WRITTEN YET, WILL NEED TO WRITE AS APPLICABLE
# # if three or more, melt down into groups, and then anova test (melt so 2 cols, 1 with group 1 with stat)
# res.aov <- aov(var ~ group, data = melted data
# summary(res.aov)
aovtab <- intable[,c(statlabel, groupvar),drop = FALSE]
# an anova is finicky and needs to have the levels be real values (which isnt surprising) - so issues will happen with blanks and NAs, and there may be times where there are blanks AND NAs, so for this case, lets change them
if (length(aovtab[,statlabel][aovtab[,statlabel] == ""]) > 0 ) {
levels(aovtab[,statlabel]) <- c(levels(aovtab[,statlabel]), "blankval")
aovtab[,statlabel][aovtab[,statlabel] == ""] <- "blankval"
}
if (length(aovtab[,statlabel][is.na(aovtab[,statlabel])]) > 0 ) {
levels(aovtab[,statlabel]) <- c(levels(aovtab[,statlabel]), "NAval")
aovtab[,statlabel][is.na(aovtab[,statlabel])] <- "NAval"
}
aovtab <- na.omit(aovtab)
## In case these values are characters, need to be factors for anova
aovtab[sapply(aovtab, is.character)] <- lapply(aovtab[sapply(aovtab, is.character)], as.factor)
aovform <- as.formula(paste0(groupvar, "~", statlabel))
aovout <- aov(aovform, aovtab)
statval <- summary(aovout)[[1]][["Pr(>F)"]][[1]]
outstatlist[[columnnum]] <- data.frame(rval = 0, pval = statval, row.names = statlabel)
}
if (class(subtab[,1]) == "numeric" | class(subtab[,1]) == "integer") {
suppressWarnings(corout <- cor.test(groupcol[,1], subtab[,1]))
outstatlist[[columnnum]] <- data.frame(rval = corout$estimate, pval = corout$p.value, row.names = statlabel)
}
}
names(outstatlist)[columnnum] <- statlabel
}
outstattable <- do.call(rbind, outstatlist)
return(outstattable)
}
# if (columnnum == 1){
# outstatlist[[groupvarnum]] <- cbind(classes = c("total", "unique_entries"), values = c(nrow(subtab), length(unique(subtab[,1]))))
# } else {
# if (class(subtab[,1]) == "character" | class(subtab[,1]) == "factor"){
# outstatlist[[groupvarnum]] <- data.frame(table(subtab[,1]))
# colnames(outstatlist[[groupvarnum]]) <- c("classes", "values")
# }
# if (class(subtab[,1]) == "numeric" | class(subtab[,1]) == "integer") {
# tempsum <- summary(subtab[,1])
# outstatlist[[groupvarnum]] <- data.frame(ids = names(tempsum), values = as.numeric(unname(tempsum)))
# colnames(outstatlist[[groupvarnum]]) <- c("classes", "values")
# }
# }
## Generate Some prestats for this analysis
### IF YOU WANT TO FIX THIS TO OUTPUT GGPLOTS - THEN USE THIS
### https://stackoverflow.com/questions/31993704/storing-ggplot-objects-in-a-list-from-within-loop-in-r
# THE FIRST COL WILL BE SAMPLE IDS TO MATCH THE FUNCTION ABOVE
# intable - the table with all the information to match from, and rownames being the samples
# groupcat - the label to split up by groups
# controlcat - the labels that are the categories to summarize
# discretevar - the sublabels of controlcat that are discrete
# contvar - the sublabels of controlcat that are continuous
summarize_table_figures <- function(intable, groupvar = NULL, outfilepath) {
dir.create(outfilepath, showWarnings = FALSE, recursive = TRUE)
## Assign our infotab as the intable either as is, or with the grouping column excluded
if (!is.null(groupvar)){
# REMOVE THE SAMPLEID COLUMN THAT IS THE FIRST COLUMN
infotab <- intable[,-1]
## Maybe I need to turn any NA in the group into an "NAgroup" so they still get counted as theyre own thing
infotab[is.na(infotab[,groupvar]),groupvar] <- "NAgroup"
subgroupvars <- as.character(unique(infotab[,groupvar]))
} else {
subgroupvars <- "ALLSAMPLES"
groupvar <- "ALLSAMPLES"
# REMOVE THE SAMPLEID COLUMN THAT IS THE FIRST COLUMN
infotab <- cbind(ALLSAMPLES = "ALLSAMPLES", intable[,-1])
}
## For each column, we have to run analysis
for (columnnum in seq_len(ncol(infotab)-1)){
catlabel <- colnames(infotab)[columnnum+1]
subtable <- infotab[,c(groupvar, catlabel)]
# If discrete - lets do stack bar charts with % and number for each group
if (class(subtable[,2]) == "character" | class(subtable[,2]) == "factor") {
meltedtab <- melt(table(subtable))
meltedtab[,2] <- as.character(meltedtab[,2])
# stack bar chart with N
pout <- ggplot(meltedtab, aes(fill = meltedtab[,catlabel], y = meltedtab[,"value"], x = meltedtab[,groupvar]))
pout <- pout + geom_bar(position="stack", stat = "identity")
pout <- pout + labs(title = catlabel, x = "category", y = "value", fill = catlabel)
pout <- pout + theme_pubr(base_size = 10, x.text.angle = 60)
pdf(paste0(outfilepath, catlabel, "_stack_barN.pdf"))
print(pout)
junk <- dev.off()
# stack bar chart with %
pout2 <- ggplot(meltedtab, aes(fill = meltedtab[,catlabel], y = meltedtab[,"value"], x = meltedtab[,groupvar]))
pout2 <- pout2 + geom_bar(position="fill", stat = "identity")
pout2 <- pout2 + labs(title = catlabel, x = "category", y = "value", fill = catlabel)
pout2 <- pout2 + theme_pubr(base_size = 10, x.text.angle = 60)
pdf(paste0(outfilepath, catlabel, "_stack_barperc.pdf"))
print(pout2)
junk <- dev.off()
}
# If continuous - lets do histograms
if (class(subtable[,2]) == "numeric" | class(subtable[,2]) == "integer") {
## Histograms plots
labsparam <- list(title = paste0(catlabel, " histogram"), x = catlabel, y = "count")
histout <- plot_histogram(data = subtable[,2,drop=FALSE], groupvar = subtable[,groupvar,drop=FALSE],
fitcurve = TRUE, labsparam, limitx = NULL, limity = NULL,
# binparam=(max(subtable[,2], na.rm = TRUE) - min(subtable[,2], na.rm = TRUE)))
binparam=min(length(unique(subtable[,2])), 20))
## Boxplots
labsparam <- list(title = paste0(catlabel, " boxplot"), x = catlabel, y = "value")
plotstatparam <- ifelse(length(unique(subtable[,groupvar])) > 1, "intra", FALSE)
bpintable <- melt(subtable, id.vars = groupvar)
bpout <- boxplot_plotter(boxplottable = bpintable, xsplit = "category", labsparam = labsparam,
testtypeparam = "t.test", colorparam = NULL, secondaxis=NULL,
plotstats = plotstatparam, comparisonparam = unique(subtable[,groupvar]))
## Write out plots
pdf(paste0(outfilepath, catlabel, "_hist.pdf"))
print(histout)
junk <- dev.off()
pdf(paste0(outfilepath, catlabel, "_bp.pdf"))
print(bpout)
junk <- dev.off()
}
}
}