-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset_assembly.R
63 lines (52 loc) · 1.71 KB
/
dataset_assembly.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
options(width=150)
# Load libraries
library(tidyverse)
# Define infiles
feature_file = "/home/johannes/proj/crse/results/2018-11-21/feba_common_genes.tab"
class_file = "/home/johannes/proj/crse/results/2018-11-22/feba_experiment_classes.tab"
fitness_file = "/home/johannes/proj/crse/data/2018-11-20/feba_genefitness.csv.gz"
# Load data
features = read_tsv(feature_file)
classes = read_tsv(class_file)
fitness = read_csv(fitness_file)
# Add missing features as NA
features = select(features, -N) %>% spread(domainId, locusId) %>%
gather(domainId, locusId, -orgId)
# Combine data
dataset = features %>% inner_join(classes) %>% left_join(fitness)
# If a feature is missing, replace its value with the average
avg_fit = filter(dataset, !is.na(fit)) %>%
group_by(domainId) %>%
summarise(fit = mean(fit))
dataset = filter(dataset, is.na(fit)) %>% select(-fit) %>%
inner_join(avg_fit) %>% rbind(filter(dataset, !is.na(fit)))
# Format data for Python scikit-learn
dataset = select(
dataset, orgId, expName, metal, antibiotics, domainId, fit
) %>% spread(domainId, fit)
feature_names = colnames(
select(dataset, -orgId, -expName, -metal, -antibiotics)
)
X_features = select(dataset, -orgId, -expName, -metal, -antibiotics)
y_metal = dataset$metal
y_antibiotics = dataset$antibiotics
# Write data files for Python scikit-learn
write(
feature_names,
"/home/johannes/proj/crse/results/2018-11-22/feba_feature_names.txt"
)
write_csv(
X_features,
"/home/johannes/proj/crse/results/2018-11-22/feba_X.csv",
col_names=F
)
write(
y_metal,
"/home/johannes/proj/crse/results/2018-11-22/feba_y_metal.txt",
ncolumns = 1
)
write(
y_antibiotics,
"/home/johannes/proj/crse/results/2018-11-22/feba_y_antibiotics.txt",
ncolumns = 1
)