snambi
/
dib-covid-dataset
forked from breno-madruga/dib-covid-dataset


  
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

	
            ##############################################################################################
################### Merging the datasets of publications related to COVID-19 #################
##############################################################################################

########################################################################
# Importing the required libraries.
import csv, re, pandas as pd, numpy as np
from string import punctuation
########################################################################

########################################################################
# 1. Defining the required functions
########################################################################

# Defining the function "clean_title".
def clean_title(title):
    if len(title) == 1 and title in punctuation:
        return None
    if title[0] in punctuation:
        title = title[1:]
    if title[-1] in punctuation:
        title = title[:-1]
    return re.sub(r"\s+", " ", title).lower()

########################################################################
# 2. Getting and preprocessing the datasets
########################################################################

########################################################################
# 2.1. arXiv
########################################################################

# Getting the data.
df_arxiv = pd.read_csv("data/prepared/arxiv_covid_19.csv", header=0,
    dtype={"id": "str"})

# Changing the type of features.
df_arxiv.loc[:, ["subject_areas", "authors"]] = df_arxiv.loc[:,
    ["subject_areas", "authors"]].apply(lambda x: x.apply(eval))
df_arxiv.publication_date = pd.to_datetime(df_arxiv.publication_date)

# Defining the feature "source".
df_arxiv["source"] = "arXiv"

# Normalizing the feature "title".
df_arxiv.title = df_arxiv.title.apply(clean_title)

########################################################################
# 2.2. bioRxiv
########################################################################

# Getting the data.
df_biorxiv = pd.read_csv("data/prepared/biorxiv_covid_19.csv", header=0,
    dtype={"id": "str"})

# Changing the type of features.
df_biorxiv.author_affil.loc[df_biorxiv.author_affil.notnull()] = df_biorxiv.author_affil.loc[
    df_biorxiv.author_affil.notnull()].apply(eval)
df_biorxiv.publication_date = pd.to_datetime(df_biorxiv.publication_date)

# Normalizing the feature "title".
df_biorxiv.title = df_biorxiv.title.apply(clean_title)

########################################################################
# 2.3. PubMed
########################################################################

# Getting the data.
df_pubmed = pd.read_csv("data/prepared/pubmed_covid_19.csv", header=0,
    dtype={"pubmed_id": "str"})

# Changing the type of features.
df_pubmed.auth_keywords.loc[df_pubmed.auth_keywords.notnull()] = df_pubmed.auth_keywords.loc[
    df_pubmed.auth_keywords.notnull()].apply(eval)
df_pubmed.author_affil.loc[df_pubmed.author_affil.notnull()] = df_pubmed.author_affil.loc[
    df_pubmed.author_affil.notnull()].apply(eval)
df_pubmed.publication_date = pd.to_datetime(df_pubmed.publication_date)

# Defining the feature "source".
df_pubmed["source"] = "PubMed"

# Normalizing the feature "title".
df_pubmed.title.loc[df_pubmed.title.notnull()] = df_pubmed.title.loc[
    df_pubmed.title.notnull()].apply(clean_title)


########################################################################
# 2.4. Scopus
########################################################################

# Getting the data.
df_scopus = pd.read_csv("data/prepared/scopus_covid_19.csv", header=0,
    dtype={"id": "str", "eid": "str", "pii": "str", "pubmed_id": "str"})

# Changing the type of features.
df_scopus.auth_keywords.loc[df_scopus.auth_keywords.notnull()] = df_scopus.auth_keywords.loc[
    df_scopus.auth_keywords.notnull()].apply(eval)
df_scopus.index_terms.loc[df_scopus.index_terms.notnull()] = df_scopus.index_terms.loc[
    df_scopus.index_terms.notnull()].apply(eval)
df_scopus.affiliations.loc[df_scopus.affiliations.notnull()] = df_scopus.affiliations.loc[
    df_scopus.affiliations.notnull()].apply(eval)
df_scopus.subject_areas.loc[df_scopus.subject_areas.notnull()] = df_scopus.subject_areas.loc[
    df_scopus.subject_areas.notnull()].apply(eval)
df_scopus.authors.loc[df_scopus.authors.notnull()] = df_scopus.authors.loc[
    df_scopus.authors.notnull()].apply(eval)
df_scopus.author_affil.loc[df_scopus.author_affil.notnull()] = df_scopus.author_affil.loc[
    df_scopus.author_affil.notnull()].apply(eval)
df_scopus.references.loc[df_scopus.references.notnull()] = df_scopus.references.loc[
    df_scopus.references.notnull()].apply(eval)
df_scopus.publication_date = pd.to_datetime(df_scopus.publication_date)

# Defining the feature "source".
df_scopus["source"] = "Scopus"

# Normalizing the feature "title".
df_scopus.title = df_scopus.title.apply(clean_title)

########################################################################
# 3. Merging/Joining the datasets
########################################################################

# Removing the duplicated records between arXiv and Scopus.
df_arxiv = df_arxiv[~df_arxiv.title.isin(df_scopus.title)]

# Removing the duplicated records between bioRxiv and Scopus.
df_biorxiv = df_biorxiv[~df_biorxiv.title.isin(df_scopus.title)]

# Removing the duplicated records between PubMed and Scopus.
idx_removed = df_pubmed.pubmed_id[df_pubmed.pubmed_id.isin(df_scopus.pubmed_id) &
                                  df_pubmed.title.isin(df_scopus.title)].index.to_list()
idx_removed += df_pubmed.pubmed_id[~df_pubmed.pubmed_id.isin(df_scopus.pubmed_id) &
                                   df_pubmed.title.isin(df_scopus.title)].index.to_list()
idx_removed += df_pubmed.pubmed_id[df_pubmed.pubmed_id.isin(df_scopus.pubmed_id) &
                                   ~df_pubmed.title.isin(df_scopus.title)].index.to_list()
df_pubmed = df_pubmed[~df_pubmed.index.isin(list(set(idx_removed)))]

# Visualizing the final number of records for each dataset.
print("arXiv:", df_arxiv.id.size)
print("bioRxiv:", df_biorxiv.id.size)
print("PubMed:", df_pubmed.pubmed_id.size)
print("Scopus:", df_scopus.id.size)
print("Expected total number of records for the final dataset:",
    (df_arxiv.id.size + df_biorxiv.id.size + df_pubmed.pubmed_id.size + df_scopus.id.size))

# Merging/Joining the datasets.
df_final = pd.concat([df_arxiv, df_biorxiv, df_pubmed, df_scopus], ignore_index=True)

# Defining the "None" value for the "NaN" values.
df_final.replace({np.nan: None}, inplace=True)

# Renaming the feature "source".
df_final.rename(columns={"source": "data_source"}, inplace=True)

# Exporting the final dataset to CSV file.
df_final.to_csv("data/raw/final_raw.csv", index=False, quoting=csv.QUOTE_ALL)