#event_type percent
event_type=df["event_type"].value_counts(
                normalize=True)

title="<b>Event Type percentage"

# Pie chart
fig = go.Figure(data=[go.Pie(
    labels=event_type.index, 
    values=event_type, 
    # textinfo = 'label+percent',
    textposition = 'inside',
    insidetextfont={'color':'white'},
    pull=[0.1, 0, 0.2,0, 0,0])])
fig.update_layout(
  title_text=title+" since 1997 Ethiopia<b>",
  titlefont={'size': 14, 'family':'Serif'},
  title_x=0.5, margin={"l":0,'r':0}
  # width=600,height=400,
)

#event_type percent
event_type <-df %>%
    group_by(event_type) %>%
    count() %>% 
    ungroup() %>% 
    mutate(perc = n / sum(n)) %>% 
    arrange(desc(perc)) %>%
    mutate(labels = scales::percent(perc))

# Pie hart
fig <- plot_ly(df, labels = ~event_type,
    type = 'pie',height=500,width=500,
    textposition = 'inside',
    insidetextfont = list(color = '#FFFFFF'),
    marker = list(line = list(
    color = '#FFFFFF', width = 1)),pull=0.03
  ) %>% layout(title= list(
  text="Event Type percentage since 1997 in Ethiopia",
  font= list(family = "Times New Roman",
  color = "black",size=15))
  )
fig

fig = px.bar(df['inter1'].value_counts()
            .sort_values(ascending=True),
            orientation='h',
            # height=400,width=800
            )
fig.update_layout(
    xaxis=axis_font,
    yaxis=axis_font,
    title_text="<b> Number of conflicts events "+
           "by Actor Type in Ethiopia (ACLED)<b>",
    titlefont={'color':'black', 'size': 16,
               'family': 'San-Serif'},
    title_x=0.5,yaxis_title=None, xaxis_title=None,
    showlegend=False,
    plot_bgcolor='white', 
    margin={"l":0,'r':0}
)

actor_type<-df %>%
    group_by(inter1) %>%
    count() %>% 
    arrange(desc(n))

p<- actor_type %>% 
  ggplot(aes(x = reorder(inter1,n), y = n)) +
  geom_bar(stat = "identity",fill="royalblue") +
  coord_flip() +
  labs(x = NULL, y = NULL)+
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank())

# With plotly
# ggplotly(p)

p

# number of event per month
number_ev=df["event_type"].groupby(
  [df['event_date'].dt.year.rename('year'), 
  df['event_date'].dt.month.rename('month')]
  ).agg({'count'})
number_ev.index = [pd.datetime(year, month,1)
          for (year,month) in number_ev.index]

fig = px.line(number_ev)

fig.update_layout(
    yaxis=axis_font,xaxis=axis_font,
    title_text="<b> Conflict events trend "+
        f"in {df['country'].unique()[0]} (ACLED)<b>",
    titlefont={'color':'black', 'size': 16,
     'family': 'San-Serif'},
    title_x=0.5,yaxis_title=None, xaxis_title=None,
    showlegend=False, plot_bgcolor='white', 
    font=dict(color ='black',),
    margin={"l":0,'r':0}
)

# number of event per month
event <- df %>% 
  mutate(year = year(event_date)
         , month = month(event_date,
          abbr = FALSE, label = TRUE)) %>% 
  group_by(year, month) %>% 
  summarise(count = n()) %>% 
  mutate(date  = as.Date(paste(year,month,1)
                    ,format="%Y %B %d"))

p <- event %>%
  ggplot( aes(x=date, y=count)) +
    geom_line(color="royalblue",size=1) +
    labs( x = NULL ,y = NULL) +
    scale_x_date(date_breaks = "5 year",
                 date_labels = "%Y")+
    theme_minimal()+
    theme(panel.grid.minor = element_blank())
p

number_fatalities = df["fatalities"].groupby(
  [df['event_date'].dt.year.rename('year'), 
   df['event_date'].dt.month.rename('month')]
   ).agg({'sum'})
number_fatalities.index = [pd.datetime(year, month,1)
       for (year,month) in number_fatalities.index]

fig = px.line(number_fatalities)
fig.update_traces(line=dict(color='red')
  ).update_layout(
    xaxis=axis_font, yaxis=axis_font,
    title_text=f"<b> Conflict fatalities trend "+
    f"in {df['country'].unique()[0]} (ACLED)<b>",
    titlefont={'color':'black', 'size': 16,
     'family': 'San-Serif'},
    title_x=0.5,yaxis_title=None, xaxis_title=None,
    showlegend=False, plot_bgcolor='white', 
    margin={"l":0,'r':0}
)

fatality <- df %>% 
  mutate(year = year(event_date), 
         month = month(event_date,
         abbr = FALSE, label = TRUE)) %>% 
  group_by(year, month) %>% 
  summarise(fat = sum(fatalities)) %>% 
  mutate(date  = as.Date(paste(year,month,1)
                   ,format="%Y %B %d"))
p <- fatality%>%
  ggplot( aes(x=date, y=fat)) +
    geom_line(color="red",size=1) +
    labs( x = NULL ,y = NULL) +
    scale_x_date(date_breaks = "5 year",
                 date_labels = "%Y")+
    theme_minimal()
p

df1 = (df[df['event_date']>'2014-12-31']
 .groupby(['year','event_type'])
 .agg({"event_date": 'count'})
  .rename(columns= {'event_date':'Count'})
 .reset_index()
)
df1['year'] = df1['year'].astype(str)

fig = px.bar(df1, x = 'year',color='event_type',
      y='Count', # height=400,width=800,
)
fig.update_layout(
    xaxis=axis_font,xaxis_title=None,
    yaxis=axis_font,yaxis_title=None,
    title_text=f"<b>Events in Ethiopia (ACLED)<b>",
    titlefont={'color':'black',
     'size': 16, 'family': 'San-Serif'},
    title_x=0.5,showlegend=True,
    legend={"font":{'size':8},'title':None},
    margin={"l":0,'r':0},
    plot_bgcolor='white', 
    font=dict(color ='black',)
)

df2 <- df %>% 
  group_by(year, event_type) %>% 
  summarise(n = n()) 


df2 %>% filter(year>2014) %>% 
  ggplot(aes(x =factor( year), y = n)) +
  geom_bar(stat = "identity",
   aes(fill = event_type)) +
  labs(x = NULL, y = NULL,fill = NULL)+
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.x = element_blank())

df2 = (df[df['event_date']>'2014-12-31']
 .groupby(['year','event_type'])
 .agg({"fatalities": 'sum'})
  # .rename(columns= {'fatalities':'sum'})/
 .reset_index()
)
df2['year'] = df1['year'].astype(str)
fig = px.bar(df2, x = 'year',color='event_type', 
    y='fatalities', # height=400,width=800,
)

fig.update_layout(
  xaxis=axis_font,xaxis_title=None,
  yaxis=axis_font,yaxis_title=None,
  title_text="<b>Fatalities in Ethiopia (ACLED)<b>",
  titlefont={'color':'black',
    'size': 16, 'family': 'San-Serif'},
  title_x=0.5,showlegend=True,
  legend={"font":{'size':8},'title':None},
  margin={"l":0,'r':0},
  plot_bgcolor='white', 
  font=dict(color ='black',)
)

df3 <- df %>% 
  group_by(year, event_type) %>% 
  summarise(fat = sum(fatalities)) 


df3 %>% filter(year>2014) %>% 
  ggplot(aes(x =factor( year), y = fat)) +
  geom_bar(stat = "identity",
   aes(fill = event_type)) +
  labs(x = NULL, y = NULL,fill = NULL)+
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.x = element_blank())

fig = px.bar(
  df['interaction'].value_counts().head(10),
    orientation='h',# height=400,width=800,
)
fig.update_layout(
  xaxis=axis_font,xaxis_title=None,
  yaxis=axis_font,yaxis_title=None,
  title_text=("<b> Most frequent Interaction"+
   " in Ethiopia conflicts (ACLED)<b>"),
  titlefont={'color':'black',
    'size': 16, 'family': 'San-Serif'},
  title_x=0.5,showlegend=False,
  legend_title=None, margin={"l":0,'r':0},
  plot_bgcolor='white', 
  font=dict(color ='black',)
)

df5<-df %>%
    group_by(interaction) %>%
    count() %>% 
    arrange(desc(n))

df5 %>% head(10)%>% 
  ggplot(aes(x = reorder(interaction,n),y=n)) +
  geom_bar(stat = "identity",fill="royalblue") +
  coord_flip() +
  labs(x = NULL, y = NULL)+
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank())

fig = px.bar(
  df['actor1'].value_counts().head(10),
  orientation='h',# height=400,width=800,
)
fig.update_layout(
  xaxis=axis_font,xaxis_title=None,
  yaxis=axis_font,yaxis_title=None,
  title_text=("<b> Most Actor involved"+
   " in Ethiopia conflicts (ACLED)<b>"),
  titlefont={'color':'black',
    'size': 16, 'family': 'San-Serif'},
  title_x=0.5,showlegend=False,
  legend_title=None, margin={"l":0,'r':0},
  plot_bgcolor='white', 
  font=dict(color ='black',)
)

df6<-df %>%
    group_by(actor1) %>%
    count() %>% 
    arrange(desc(n))

df6 %>% head(10)%>% 
  ggplot(aes(x = reorder(actor1, n), y = n)) +
  geom_bar(stat = "identity",fill="royalblue") +
  coord_flip() +
  labs(x = NULL, y = NULL)+
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank(),
        axis.text.y = element_text(size=9))

# ? we change the theme to dark
fig = px.bar(
  df['actor1'].value_counts().head(10),
     orientation='h',# height=400,width=800,
)
fig.update_layout(
  xaxis_title=None,yaxis_title=None,
  title_text=("<b> Most Actor involved"+
   " in Ethiopia conflicts (ACLED)<b>"),
  titlefont={'color':'red',
    'size': 16, 'family': 'San-Serif'},
  title_x=0.5,showlegend=False,
  legend_title=None, margin={"l":0,'r':0},
  template="plotly_dark",
)

df6 %>% head(10)%>% 
  ggplot(aes(x = reorder(actor1, n), y = n)) +
  geom_bar(stat = "identity",fill="royalblue") +
  coord_flip() +
  labs(x =NULL, y = NULL)+
  theme_ft_rc()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank(),
        axis.text.y = element_text(size=9))

df5 = df['actor1'].value_counts()\
    .rename('count').reset_index().head(10)
sns.set(rc={'axes.facecolor':'#FFFFFF', 
    'figure.facecolor':'#FFFFFF',
    'font.sans-serif':'Arial',
    'font.family':'sans-serif'
})
fig,ax = plt.subplots(figsize=(10,6))

ax = sns.barplot(y="index", x= "count",data=df5,
      edgecolor="black",palette=colors, hatch='/')
for p in ax.patches:
    value = f'{p.get_width():g}'
    x = p.get_x() + p.get_width() + 100
    y = p.get_y() + p.get_height() / 2 
    ax.text(x, y, value, ha='left', va='center', 
      bbox=dict(facecolor='none', edgecolor='black',
                boxstyle='round', linewidth=1))
ax.set(xticklabels=[],ylabel=None, xlabel=None,);
plt.tight_layout(pad=1, w_pad=1, h_pad=1)
ax

df6 %>% head(10)%>% 
  ggplot(aes(x = reorder(actor1, n), y = n)) +
  geom_bar(stat = "identity",aes(fill = actor1),
   show.legend = FALSE) +
  coord_flip() +
  labs(x = NULL, y = NULL)+
  geom_label(aes(label = formatC(n, format="f"
    ,big.mark=",", digits=0)), size = 2.2) +
  theme_minimal()+
  theme(panel.grid.minor = element_blank(),
        panel.grid.major.y = element_blank(),
        axis.text.y = element_text(size=9))

from os import path
from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
import os

from wordcloud import WordCloud, STOPWORDS

l=[]
for x in list(df['notes'].values):
    if str(x) not in l:l.append(str(x))

text = ' '.join(l)

stopwords = set(STOPWORDS)
for i in ['size','no','report','said',"ongoing",
    'arrested','reported','area','several']:
    stopwords.add(i)

wc = WordCloud(background_color="white",
  max_words=2000,
  # mask=None,
  colormap ='Dark2',
  stopwords=stopwords,
  # contour_width=3, contour_color='steelblue'
)

# generate word cloud
wc.generate(text.lower());

plt.figure()
# Show the wordcloud
plt.imshow(wc, interpolation='bilinear')
plt.axis("off");
plt.show()

# # Install
# install.packages("tm")  # for text mining
# install.packages("SnowballC") # for text stemming
# install.packages("wordcloud") # word-cloud generator 
# install.packages("wordcloud2")

# install.packages("RColorBrewer") # color palettes
# Load
library("tm")
library("SnowballC")
library("wordcloud")
library("wordcloud2")
library("RColorBrewer")

#Create a vector containing only the text
text <- df$notes
# Create a corpus  
docs <- Corpus(VectorSource(text))

# text transformation
toSpace <- content_transformer(
  function (x , pattern ) gsub(pattern, " ", x))
docs <- tm_map(docs, toSpace, "/")
docs <- tm_map(docs, toSpace, "@")
docs <- tm_map(docs, toSpace, "\\|")


# Convert the text to lower case
docs <- tm_map(docs, content_transformer(tolower))
# Remove numbers
docs <- tm_map(docs, removeNumbers)
# Remove english common stopwords
docs <- tm_map(docs,removeWords,stopwords("english"))
# Remove your own stop word
# specify your stopwords as a character vector
docs <- tm_map(docs, removeWords,
 c('size','no','report','said',
'arrested','reported','area','several')) 
# Remove punctuations
docs <- tm_map(docs, removePunctuation)
# Eliminate extra white spaces
docs <- tm_map(docs, stripWhitespace)

dtm <- TermDocumentMatrix(docs)
matrix <- as.matrix(dtm)
words <- sort(rowSums(matrix),decreasing=TRUE)
dfw <- data.frame(word = names(words),freq=words)

set.seed(1234) # for reproducibility 

# wordcloud2(data=dfw, size=1.6, color='Dark2')

wordcloud(words = dfw$word, freq = dfw$freq,
 min.freq = 50, max.words=2000,
  random.order=FALSE,colors=brewer.pal(8, "Dark2"))