The structure of recent philosophy (II)

1
2
3
4

date_string = datetime.datetime.now().strftime("%Y-%m-%d-%H:%M")


RC = mk.RecordCollection("datanew")

 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

# print(RC.glimpse())
RC2 = mk.RecordCollection()

for R in RC:
    randnr = randint(0, 4)
    
    if len(R.getCitations().get("author"))>=1: # and randnr==0 apply condition in order to downsample records
        #Here we kick out every paper that cites less then 3 authors. Why? because they
        #are so dissimilar from the others, that they only produce noise.  
   
        try:
            R['year']
            #if R['year']>1961:
            #R['abstract']  #Add this when working with abstracts. It removes every paper that has none. 
            #This can sometimes remove whole journals, that are archived without abstracts, so handle with care.
            RC2.add(R)
        except KeyError:
            pass
    else:
        pass
    

print(RC2.glimpse())


RC = RC2

 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

########### Cited Works - Features ############

from sklearn.feature_extraction.text import TfidfVectorizer, CountVectorizer
import re
drc = pd.DataFrame.from_dict(RC.forNLP(extraColumns=['journal','AU','FU','PD']))

d = []
citedAU = []
citestring =[]
for R in RC:

    d.append(list(set(R.getCitations().get("citeString")))) #To cluster by cited author
    citedAU.append(list(set(R.getCitations().get("author"))))
    citestring.append(list(set(R.getCitations().get("citeString"))))

drc["citedAU"] = citedAU
drc["citestring"] = citestring
#print(d[0])
authorslist = ['§'.join(filter(None,x)) for x in list(d)] 
#print(authorslist[0])

# vec = TfidfVectorizer(token_pattern=r'(?<=[^|§])[\s\w,\.:;]+(?=[$|§])')
vec = CountVectorizer(token_pattern=r'(?<=[§])[\s\w,\.:;\/\[\]-]+(?=[§])',binary=True, min_df = 3)#, min_df = 1)


Xrc = vec.fit_transform(authorslist)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17

########### Authors - Features ############

d = []
for R in RC:
    authors = list(set(R.getCitations().get("author")))
#    print(authors)
    authors = filter(None, authors)
    f = []
    for a in authors:
        f.append(' '.join([w for w in a.split(' ')if len(w)>2]))
        
    authors = f#' '.join(f)
    d.append(authors)
authorslist = [';'.join(filter(None,x)) for x in list(d)] 
vec = CountVectorizer(token_pattern=r'(?<=[;])[\s\w]+(?=[;])',binary=True, min_df = 10)

XrcAu = vec.fit_transform(authorslist)

1
2

k = [XrcAu,Xrc]
XrcFull = scipy.sparse.hstack(k).tocsr()

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21

###### Filtering #######
from scipy.sparse import coo_matrix, vstack
from scipy.sparse import csr_matrix
import scipy as scipy
row_names = np.array(drc["id"])

newdf=[]
a = 0
# index by name:
for x in range(0,XrcFull.shape[0]): #Xrc.shape[0]):
    row_idx, = np.where(row_names == drc["id"][x])
    if np.diff(XrcFull[row_idx].tocsr().indptr) >= 4:
        if a == 0:
            k = [XrcFull[row_idx]]
        if a != 0:
            k.append(XrcFull[row_idx])
        a = a+1
        newdf.append(drc.loc[x])
        
drc = pd.DataFrame(newdf).reset_index()
M = scipy.sparse.vstack((k))

1
2
3
4
5
6
7
8
9

from sklearn.decomposition import TruncatedSVD
SVD = TruncatedSVD(n_components=350, n_iter=7, random_state=42)

XSVD = SVD.fit_transform(M)
print(SVD.explained_variance_ratio_.sum())
dSVD = pd.DataFrame(XSVD)

sSVD = dSVD[[0,1]]
sSVD.columns = ['x','y']

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

import umap

try:
    drc = drc.drop('x',axis=1)
    drc = drc.drop('y',axis=1)

except KeyError:
    pass


embedding = umap.UMAP(n_neighbors = 7,#small => local, large => global: 5-50
                      min_dist = 0.0005, #small => local, large => global: 0.001-0.5
                      spread = 1.5,
                      metric='cosine').fit_transform(XSVD)
embedding = pd.DataFrame(embedding)
embedding.columns = ['x','y']

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19

%%R -i embedding --width 1200 --height 800 -r 140 --bg #F5F5F5
library(hrbrthemes)
library(ggplot2)
library(fields)
embedding$density <- fields::interp.surface(
  MASS::kde2d(embedding$x, embedding$y), embedding[,c("x","y")])

p <- ggplot(embedding, aes(x=embedding$x, y=embedding$y,alpha = 1/density))+

 guides(alpha=FALSE)+

geom_point(color="#3366cc", pch=16,cex=1.2)+ theme_ipsum_rc()+
labs(x="", y="",
       title="The 2d-reduction by UMAP",
       subtitle="...based on the code by McInnes, Healy (2018)",
       caption="by Maximilian Noichl")+
theme(panel.grid.major = element_line(colour = "lightgrey"),panel.grid.minor = element_blank()
)
p

1

drc['timestamp'] = drc["year"] + drc["PD"].fillna(0).replace('', 0, regex=True)/12

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

n_neighbors = 7
embeddingI = umap.UMAP(n_components=1,
                        n_neighbors=n_neighbors,
                      min_dist=0.008,
                       spread = 0.8,
                      metric='cosine').fit(XSVD)

coordinates = []
for year in range(int(drc["year"].min()),int(drc["year"].max())):
    l = list(np.where(drc["year"] == year)[0])
    L = XSVD[l]     

    emb = embeddingI.transform(L)
    emb = pd.DataFrame(emb)
    emb.columns = ['xI']

    emb["year"] = drc.iloc[l,:]["timestamp"].tolist()
    coordinates.append(emb)

coordinates = pd.concat(coordinates, ignore_index=True)

embeddingI = pd.DataFrame(embeddingI.embedding_)
embeddingI.columns = ['xI']
embeddingI["year"] = drc["timestamp"].tolist()

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19

%%R -i coordinates --width 1200 --height 800 -r 140 --bg #F5F5F5
library(hrbrthemes)
library(ggplot2)
library(fields)
coordinates$density <- fields::interp.surface(
  MASS::kde2d(coordinates$xI, coordinates$year), coordinates[,c("xI","year")])

p <- ggplot(coordinates, aes(x=coordinates$xI, y=coordinates$year,alpha =1/density))+#1/density

 guides(alpha=FALSE)+

geom_point(color="#3366cc", pch=16,cex=1.2)+ theme_ipsum_rc()+
labs(x="", y="",
       title="The temporal UMAP-embedding",
       subtitle="...based on the code by McInnes, Healy (2018)",
       caption="by Maximilian Noichl")+
theme(panel.grid.major = element_line(colour = "lightgrey"),panel.grid.minor = element_blank()
)
p

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

try:
    drc = drc.drop('cluster',axis=1)
except KeyError:
    pass

import hdbscan

#(min_cluster_size=500, min_samples=30, gen_min_span_tree=True)
#clusterer = hdbscan.HDBSCAN(min_cluster_size=455, min_samples=35, gen_min_span_tree=True)

clusterer = hdbscan.HDBSCAN(min_cluster_size=440, min_samples=50, gen_min_span_tree=True)
clusterer.fit(embedding)
XCLUST = clusterer.labels_
clusternum = len(set( clusterer.labels_))-1


dfclust = pd.DataFrame(XCLUST)
dfclust.columns = ['cluster']


print(clusternum)
### Let's play a little sound when we're done:
import winsound
winsound.Beep(550,300)

 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

%%R -i embedding,dfclust,embeddingI -o myNewColors,tmeans --width 1200 --height 1800 -r 140 --bg #F8F4E9

library(hrbrthemes)
library(ggplot2)
library(fields)
library(plyr)
options(warn=0)# 0 zum anschalten

#Get the cluster means:
means <- aggregate(embedding[,c("x","y")], list(dfclust$cluster), median)
means <- data.frame(means) 
n=nrow(means)
means <- means[-1,]

#Make the colors: 
mycolors <- c("#c03728","#919c4c","#fd8f24","#f5c04a","#e68c7c","#00666b","#142948","#6f5438") 

pal <- colorRampPalette(sample(mycolors))
s <- n-1
myGray <- c('#95a5a6')
myNewColors <- sample(pal(s))
myPal <- append(myGray,myNewColors)

#get temporal means:
tmeans <- aggregate(embeddingI[,c("xI","year")], list(dfclust$cluster), median)
tmeans <- data.frame(tmeans) 
tmeans <- tmeans[-1,]




#get density, to avoid overplotting
embedding$density <- fields::interp.surface(
  MASS::kde2d(embedding$x, embedding$y), embedding[,c("x","y")])

#get temporal density
embeddingI$density <- fields::interp.surface(
  MASS::kde2d(embeddingI$xI, embeddingI$year), embeddingI[,c("xI","year")])


p <- ggplot(embedding, aes(x=embedding$x, y=embedding$y, color= as.factor(dfclust$cluster), alpha = 1/density))+
geom_point(pch=20,cex=1.6)+ 
theme_ipsum_rc()+
scale_color_manual(values = myPal) +
 guides(alpha=FALSE, color=FALSE)+
geom_point(data=means, aes(x=means$x, y=means$y), color= myNewColors, alpha = 1,size =6)+
annotate("text", x = means[,c("x")], y = means[,c("y")], label = means[,c("Group.1")], color="white", fontface="bold",  size=3.2, parse = TRUE, hjust=0.5)+
labs(x="", y="",
       title="The clusters found by hdbscan...",
       subtitle="a density-based clustering algorithm. Embedded with UMAP in two dimensions...")+
theme(panel.grid.major = element_line(colour = "lightgrey"),panel.grid.minor = element_blank())


t <- ggplot(embeddingI, aes(x=embeddingI$xI, y=embeddingI$year, color= as.factor(dfclust$cluster), alpha = 1/density))+
geom_point(pch=20,cex=1.2)+
theme_ipsum_rc()+
scale_color_manual(values = myPal) +
guides(alpha=FALSE, color=FALSE)+
geom_point(data=tmeans, aes(x=tmeans$x, y=tmeans$y), color= myNewColors, alpha = 1,size =4.2)+
annotate("text", x = tmeans[,c("xI")], y = tmeans[,c("year")], label = tmeans[,c("Group.1")], color="white", fontface="bold",  size=2.2, parse = TRUE, hjust=0.5)+
labs(x="", y="Publication date",
         subtitle="...and one dimension, overlayed with publication dates on the y-axis.",
       caption="by Maximilian Noichl")+
theme(panel.grid.major = element_line(colour = "lightgrey"),panel.grid.minor = element_blank())


library(gridExtra)
grid.arrange(p,t, ncol = 1,  heights = c(1, 1))

# pdf("ClusteringUMap.pdf", width = 12, height = 12) # Open a new pdf file
# grid.arrange(p,t, ncol = 1,  heights = c(1, 1)) # Write the grid.arrange in the file\n",
# dev.off()

1
2
3
4
5
6

import matplotlib.colors
plt.rcParams['figure.figsize'] = [20, 20]
cmap = matplotlib.colors.LinearSegmentedColormap.from_list("", ["#bfc4c6","#bfc4c6","#bfc4c6"])

clusterer.condensed_tree_.plot(cmap = cmap,select_clusters=True,label_clusters = True, selection_palette=myNewColors, 
                               colorbar = False,max_rectangles_per_icicle=80, alpha=0.7,barwidthfactor=5,linecolor='#bfc4c6',linewidth=2)

1
2
3

drc = pd.concat([drc, dfclust],axis=1)
drc = drc.dropna(subset=['cluster'])
drc = pd.concat([drc, embedding],axis=1)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

fullstrsl = []
for x in range(0,clusternum):
    abstracts = list(drc.loc[drc['cluster'] == x]['abstract'])
    abstracts = ";".join(str(x) for x in abstracts).replace('|',' ').replace('paper',' ').replace('argue',' ').replace('account',' ').replace('theory',' ') #kick out common abstract words of no importance
    fullstrsl.append(abstracts)
    
vec = CountVectorizer( stop_words='english')#Choose CountVectorizer for the most common words in the cluster, TfidfVectorizer for the words with the greatest differentiation value.
X = vec.fit_transform(fullstrsl)
#print(pd.DataFrame(X.toarray(), columns=vec.get_feature_names())) #To look into the vectors. Beware, can take a bit of RAM


clusterfeatures = pd.DataFrame(X.toarray(), columns=vec.get_feature_names())
fullscore = []
for x in range(0,clusternum):
    scores = zip(vec.get_feature_names(), np.asarray(X[x,:].sum(axis=0)).ravel())
    sorted_scores = sorted(scores, key=lambda x: x[1], reverse=True)
    myscores = sorted_scores[0:20]
    
    scorelist = []
    for s in myscores:
        scorelist.append(s[0])
    fullscore.append(scorelist)
display(pd.DataFrame(fullscore).transpose())

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

fullstrsl = []
for x in range(0,clusternum):
    abstracts = list(drc.loc[drc['cluster'] == x]['abstract'])
    abstracts = ";".join(str(x) for x in abstracts).replace('|',' ').replace('paper',' ').replace('reserved',' ').replace('argue',' ').replace('account',' ').replace('theory',' ')
    fullstrsl.append(abstracts)
    
vec = CountVectorizer( stop_words='english', ngram_range=(2, 2))#Choose CountVectorizer for the most common words in the cluster, TfidfVectorizer for the words with the greatest differentiation value.
X = vec.fit_transform(fullstrsl)
#print(pd.DataFrame(X.toarray(), columns=vec.get_feature_names())) #To look into the vectors. Beware, can take a bit of RAM


clusterfeatures = pd.DataFrame(X.toarray(), columns=vec.get_feature_names())
fullscore = []
for x in range(0,clusternum):
    scores = zip(vec.get_feature_names(), np.asarray(X[x,:].sum(axis=0)).ravel())
    sorted_scores = sorted(scores, key=lambda x: x[1], reverse=True)
    myscores = sorted_scores[0:20]
    
    scorelist = []
    for s in myscores:
        scorelist.append(s[0])
    fullscore.append(scorelist)
display(pd.DataFrame(fullscore).transpose())

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

fullstrsl = []
for x in range(0,clusternum):
    authors = list(drc.loc[drc['cluster'] == x]['citedAU'])
    authors = [item for sublist in authors for item in sublist]
    authors = " §".join(str(x) for x in authors)
    authors = ' '.join( [w for w in authors.split() if len(w)>2] )
    fullstrsl.append(authors)

#print(fullstrsl[1])
vec = CountVectorizer(token_pattern=r'[\s\w\.-]+(?=[$|§])')#Choose CountVectorizer for the most common words in the cluster, TfidfVectorizer for the words with the greatest differentiation value.
X = vec.fit_transform(fullstrsl)

clusterfeatures = pd.DataFrame(X.toarray(), columns=vec.get_feature_names())
fullscoreA = []
for x in range(0,clusternum):
    scores = zip(vec.get_feature_names(), np.asarray(X[x,:].sum(axis=0)).ravel())
    sorted_scores = sorted(scores, key=lambda x: x[1], reverse=True)
    myscores = sorted_scores[0:10]
    
    scorelist = []
    for s in myscores:
        scorelist.append(s[0])
    fullscoreA.append(scorelist)
display(pd.DataFrame(fullscoreA).transpose())

  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

%%R -i embedding,dfclust,myNewColors -o labelpol,cltest --bg #fbf8f1
#-h 1600 -w 1600 -r 140 --bg #fbf8f1

# Some imports:
library(hrbrthemes)
library(ggplot2)
library(fields)
# library(ggrepel)
library(ggforce)
#install.packages('ggalt')
library(ggalt)
library(stringr)



options(warn=0)# 0 zum anschalten

#Get the cluster means:
means <- aggregate(embedding[,c("x","y")], list(dfclust$cluster), mean)
means <- data.frame(means)
#And Variance, for the labels:
test <- aggregate(embedding[,c("x")], list(dfclust$cluster), var)
test <- test[-1,]

n=nrow(means)
means <- means[-1,]
# #Make the colors: 
# mycolors <- c("#dd593c",
# "#ead96c",
# "#df4467",
# "#8d2315",
# "#675d69",
# "#70897b",
# "#131541") 

# pal <- colorRampPalette(sample(mycolors))
# s <- n-1
myGray <- c('#95a5a6')
# myNewColors <- sample(pal(s))
myPal <- append(myGray,myNewColors)

#fonts:
# library(showtext)
# font.add.google(name = "Alegreya Sans SC", family = "SC")
# showtext.auto()


#◇

labels <- c("Responsibility & Free Will","Classical Logic & Paradoxes","Formal Th. of Science & Modeling","Ph. of linguistics & semantics",
            "Formal Epistemology","Causality & Scient. Explanation","Probability & Decision Theory","History of Science",
            "Possible Worlds & Counterfactuals",
            "Quantum Mechanics","Epistemology: Foundationalism","Epistemology: Contextualism","Biology & Evolution",
            "Virtue Ethics","Practical Reason","Liberalism & Legal Ph.","Moral Judgement","Art & Aesthetics","Cognitive Science",
            "Ph. of Mind: Phenomenology","Ph. of Mind: Mental Causation","Possible Worlds: Metaphysics","Reference, Proper Names",
            "Reference of Math. Objects","Quinean Ontology")
labelsb <- c("J. M. Fischer, P. van Inwagen","","","",
            "","N. Cartwright, J. Woodward","Dutch Book, Sleeping Beauty,...","Newton",
            "",
            "","","","",
            "Aristotle","Kant & Hume","","","","",
            "","","","",
            "","")
#circular markers:  
library(gridExtra)

circle <- polygon_regular(100)
pointy_points <- function(x, y, size){
  do.call(rbind, mapply(function(x,y,size,id) 
    data.frame(x=size*circle[,1]+x, y=size*circle[,2]+y, id=id),
         x=x,y=y, size=size, id=seq_along(x), SIMPLIFY=FALSE))
}


#get density, to avoid overplotting
embedding$density <- 1/ as.numeric(fields::interp.surface(MASS::kde2d(embedding$x, embedding$y), embedding[,c("x","y")]))

                        
# get for every label, wether it is in the + or - part of the x-axis:                        
xpol <-abs(means[,c("x")])/means[,c("x")]
ypol <-abs(means[,c("y")])/means[,c("y")]


polfact <- 1.5
 
# build a circle for the labels:
r <- 14.5
sequence <- seq(from = 1, to = s, by = 1)
angles <- 360/s*sequence
angle <-(angles*(pi/180))
         
xlabl <- cos(angle)*r
ylabl <- sin(angle)*r
circlecord <- cbind(xlabl,ylabl)

labelpol <-abs(circlecord[,c("xlabl")])/circlecord[,c("xlabl")]




# install.packages("stringr")
# library(stringr)



coord_x=5
coord_y=5
                        
# define circular markers:
circular_annotations <- pointy_points(means$x, means$y, size=test$x*0.25+1)
embedding <- cbind(embedding,dfclust)
filtered <- as.data.frame(subset(embedding, cluster >= 0))
cltest <- filtered
                        
                        
#Let's plot!
p <- ggplot(data=filtered, aes(x=x, y=y, color= as.factor(cluster), alpha='density'))+
geom_point(pch=20,cex=1.5)+#, alpha = 0/density)#+ 
#scale_color_manual(values = myPal)+
scale_x_continuous(limits=c(-18.5,18.5))+
scale_y_continuous(limits=c(-15,15))
#geom_polygon(data=circular_annotations, aes(x,y,group=factor(id), fill = factor(id)),alpha=0.15)+
#scale_fill_manual(values = myNewColors) +     
                       
# guides(alpha=FALSE, color=FALSE, fill=FALSE)+


q <- ggplot_build(p + stat_density2d(n=800,h=c(1.1,1.1)))$data[[2]]
q <- q[str_detect(q$group, "001") == TRUE, ]
# z <- max(test$x)
# print(z)

o <- aggregate(q$x, list(q$group) , min)
z <- subset(q,subset = q$x %in%  c(o$x))

omax <- aggregate(q$x, list(q$group) , max)
zmax <- subset(q,subset = q$x %in%  c(omax$x))

c <- data.frame()
count <- 1
for (val in xpol) {
if(val <0 ){
    c <- rbind(c, z[count,])
} else {
    c <- rbind(c, zmax[count,])
}
    count <- count + 1
}

contactpoints <-  data.frame(c$x,c$y)

#Append every label to its best fit on that circle, using the hungarian algorithm:
require(clue)

distances <- rdist(circlecord,contactpoints)
sol <- solve_LSAP(t(distances))
solo <- data.frame(cbind(mx=(contactpoints[,1]), my=(contactpoints[,2]), cx=(circlecord[sol, 1]), cy=(circlecord[sol, 2])))
                        
xcpol <-abs(solo$cx)/solo$cx



r <- 
p +
geom_point(data=subset(embedding, cluster == -1), aes(x=x, y=y),pch=20,cex=1.5,alpha=0.2, color=myGray)+                      
geom_polygon(data=q, aes(x,y ,group = as.factor(q$group),fill = as.factor(q$group)),color= NA,alpha=0.3,linetype=1,size=0.6)+ # ,linetype=3, color="black"
#scale_fill_manual(values = myPal)+
theme_ipsum()+
# scale_fill_manual(values = myPal)
guides(alpha=FALSE, color=FALSE, fill=FALSE)+
geom_point(data=solo, aes(x=cx, y=cy), color= myNewColors, alpha = 1,pch=16,size=3, stroke = 1)+                      

labs(x="UMAP-x", y="UMAP-y",
    title="The structure of recent Philosophy",
    subtitle="A umap & hdbscan-cluster-analysis of ~ 50000 papers in philosophy that brings out the major groupings of the discipline.",
    caption="by Maximilian Noichl, 2018")+               
theme(panel.grid.major = element_line(colour = "grey", linetype="dotted", size=0.55),panel.grid.minor = element_blank())+
theme(plot.background = element_rect(fill = "#fbf8f1"))+
#expand_limits(x = c(r+10,0-r-10),y = c(r+10,0-r-10))+
#theme(plot.title = element_text(size=27, family="SC", face="plain"))+
coord_fixed()+
annotate("segment", x = c$x, y = c$y, xend = c$x+xcpol*0.5, yend =c$y, color= myNewColors,alpha=0.3, size = 1)+
annotate("segment", x = c$x+xcpol*0.5, y =c$y, xend = solo$cx, yend = solo$cy, color=myNewColors,alpha=0.3, size = 1)+
# theme(plot.margin=grid::unit(c(0,0,0,0), "mm"))+

#annotate("segment", x = (solo$mx+(test$x*0.25+1)*xcpol)+xcpol*0.3, y = solo$my, xend = solo$cx, yend = solo$cy, color= myNewColors, alpha = 0.25, size = 0.7)+        
annotate("text", x = solo$cx+xcpol*0.3, y = solo$cy+0.19, parse = FALSE, label = labels, color="black", fontface="bold", family = "sans", size=3, hjust=abs((xcpol+1)/2-1),vjust=1)+
annotate("text", x = solo$cx+xcpol*0.3, y = solo$cy+0.3-0.5, parse = FALSE, label = labelsb, color="black", fontface="italic", family = "sans", size=3, hjust=abs((xcpol+1)/2-1),vjust=1)+             
#geom_point(data=c, aes(x,y),color="black")+
scale_color_manual(values = myNewColors)+
scale_fill_manual(values = myNewColors)+
theme(plot.background=element_rect(fill=NA, colour=NA))+



NULL

r
ggsave('plot.png', plot = last_plot(),width=15,height=15,dpi=300)