From a3526a71507f762d36b205275230d73b5e94cf18 Mon Sep 17 00:00:00 2001
From: ChayneAzriel <61120310+ChayneAzriel@users.noreply.github.com>
Date: Tue, 24 Nov 2020 09:12:02 +0200
Subject: [PATCH] Create case_study_250020.ipynb

---
 case_study_250020.ipynb | 1047 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 1047 insertions(+)
 create mode 100644 case_study_250020.ipynb

diff --git a/case_study_250020.ipynb b/case_study_250020.ipynb
new file mode 100644
index 0000000..6fd657e
--- /dev/null
+++ b/case_study_250020.ipynb
@@ -0,0 +1,1047 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import numpy as np
+import scipy
+from scipy import stats
+from scipy.stats import chi2
+
+#importing data
+file_path = 'involved_markers_hebrew.csv'
+involved_markers_hebrew = pd.read_csv(file_path)
+data = involved_markers_hebrew
+
+#data functions
+def filtered_data(data, column):
+    data_filtered = data.copy()
+    data_filtered.sort_values(column)
+    data_filtered.drop_duplicates(subset = column, keep = 'first', inplace = True)
+    return data_filtered
+    
+def road_section(data, section_no):
+    road_section = (data['road_segment_id']) == int(section_no)
+    data_section = data[road_section]
+    return data_section
+def road_type(data,type_no):
+    type_no = (data['road_type']) == int(type_no)
+    data_road = data[type_no]
+    return data_road
+def road_street(street_no):
+    road_street = (data['street1']) == int(street_no)
+    data_street = data[road_street]
+    return data_street
+
+def data_years(data,start_year, end_year):
+    data_years = (data['accident_year'] >= int(start_year)) & (data['accident_year'] <= int(end_year))
+    return data[data_years]
+
+def data_fatal(data):
+    data_fatal = data[data['accident_severity'] == 1]
+    return data_fatal
+
+def fatal_road_section(section_no):
+    road_section = (data_fatal['road_segment_id']) == int(section_no)
+    data_fatal_section = data_fatal[road_section]
+    return data_fatal_section
+    
+  #creating different values for variables
+
+data['involve_vehicle_type_t'] = data['involve_vehicle_type'].replace(to_replace = {1:1,2:2,3:3,4:4,5:4,6:4,7:4,8:5,9:5,10:5,11:5,12:1,13:5,14:5,15:5,16:5,17:5,18:5,19:5,20:5,21:5,22:5,23:5,24:4,25:4})
+
+'''1==Private car, 2==Truck, 3==Motorcycle, 4==Bus, 5==Cab, 6==Other'''
+data['involve_vehicle_type'] = data['involve_vehicle_type'].replace(to_replace = {1:1,2:2,3:2,4:2,5:2,6:2,7:2,8:3,9:3,10:3,11:4,12:5,13:6,14:6,15:6,16:6,17:6,18:4,19:3,20:3,21:10,22:6,23:6,24:2,25:2})
+
+'''1==day light, 2==night light, 3==night no sight, 4==night, 5==twilight'''
+data['road_light'] = data['road_light'].replace(to_replace = {1:1,2:3,3:2,4:3,5:3,6:4,7:3,8:3,9:3,10:5,11:1})
+
+data['injury_severity'] =data['injury_severity'].replace(to_replace={1:3,2:2,3:1}).astype('category')
+
+data['accident_severity'] =data['accident_severity'].replace(to_replace={1:3,2:2,3:1}).astype('category')
+
+data['medical_type'] = data['medical_type'].replace(to_replace={9:0}).astype('category')
+
+data['speed_limit'] =data['speed_limit'].replace(to_replace={0:0, 1:50, 2:60, 3:70, 4:80, 5:90, 6:100, 7:110, 8:120})
+
+data['road_width'] = data['road_width'].replace(to_replace={0:0, 1:5, 2:7, 3:10.5, 4:14, 5:20})
+
+data['engine_volume_hebrew'] = data['engine_volume_hebrew'].astype('category')
+
+#creating unique data for markers analysis (reduces from 1897266 to 771638 [duplicates = 1125628])
+data_filtered = filtered_data(data, 'provider_and_id')
+
+##datat for accident severity harsh & fatal
+data_fatal = data_fatal(data)
+data_filtered_fatal = filtered_data(data_fatal, 'provider_and_id')
+
+#data different types of the road
+data_4 = road_type(data, 4)
+data_3 = road_type(data, 3)
+
+data_filtered_4 = road_type(data_filtered, 4)
+data_filtered_3 = road_type(data_filtered, 3)
+
+#create different data frames by road section and road type for involve
+#road_segment_name: צומת דימונה - צומת בית אשל
+data_4_250020 = road_section(data_4, 250020)
+
+road_section_drop_4 = data_4.road_segment_id != (250020)
+data_4 = data_4[road_section_drop_4]
+
+data_3_250020 = road_section(data_3, 250020)
+
+road_section_drop_3 = data_3.road_segment_id != (250020)
+data_3 = data_3[road_section_drop_3]
+
+data_250020 = road_section(data, 250020)
+
+road_section_drop = data.road_segment_id != (250020)
+data = data[road_section_drop]
+
+#create different data filtered frames by road section and road type for markers
+data_filtered_4_250020 = road_section(data_filtered_4, 250020)
+
+road_section_drop_4 = data_filtered_4.road_segment_id != (250020)
+data_filtered_4 = data_filtered_4[road_section_drop_4]
+
+data_filtered_3_250020 = road_section(data_filtered_3, 250020)
+
+road_section_drop_3 = data_filtered_3.road_segment_id != (250020)
+data_filtered_3 = data_filtered_3[road_section_drop_3]
+
+df1 = data_filtered_4_250020.groupby(['accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+# Data to plot
+labels =  'minor','harsh','fatal'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0]]
+colors = ['#f9f998','#f9ac86','#fc999c']
+#cfac86
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident severity on road 25", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.1.1.png')
+plt.show()
+
+df1 = data_filtered_4.groupby(['accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['4']
+# Data to plot
+labels =  'minor','harsh','fatal'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0]]
+colors = ['#f9f998','#f9ac86','#fc999c']
+#cfac86
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident severity on other interurban roads", loc='left', fontsize=10.5, fontweight=0, color='k')
+#plt.savefig(r'slide_25.1.2.png')
+plt.show()
+
+df1 = data_4_250020.groupby(['involve_vehicle_type']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+df
+
+height = [df.iloc[0,0], df.iloc[1,0], df.iloc[3,0], df.iloc[4,0], df.iloc[2,0],   df.iloc[5,0]]
+bars = ('private','truck','bus', 'cab','motorcycle','other')
+y_pos = np.arange(len(bars))
+ 
+# Create horizontal bars
+plt.barh(y_pos, height, color = '#f9ac86')
+ 
+# Create names on the y-axis
+plt.yticks(y_pos, bars)
+
+
+y = [0,1,2,3,4,5]
+label = [916, 250,105,21,9, 68]
+for i in range(0,6):
+    plt.text(x = label[i]+2 , y = y[i], s = label[i], size = 11, color = 'k')
+
+#title
+plt.title("involve vehicle type", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+plt.box(on=None)
+
+#plt.savefig(r'slide_25.2.png')
+
+# Show graphic
+plt.show()
+
+df1 = data_4_250020.groupby(['population_type']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+df.apply(lambda x: x/x.sum()).round(2)
+height = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0]]
+bars = ('jewish', 'arab', 'other', 'forien')
+y_pos = np.arange(len(bars))
+ 
+# Create horizontal bars
+plt.barh(y_pos, height, color = '#f9f998')
+ 
+# Create names on the y-axis
+plt.yticks(y_pos, bars)
+
+x = 680
+y = [0,1,2,3]
+label = ['51%', '46%', '2%', '1%']
+for i in range(0,4):
+    plt.text(x = x , y = y[i], s = label[i], size = 11, color = 'k')
+
+#title
+plt.title("road 25 population type", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+plt.box(on=None)
+
+#plt.savefig(r'slide_25.3.png')
+
+# Show graphic
+plt.show()
+
+df1 = data_4_250020.groupby(['accident_year', 'injury_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['year','severity','count']
+df = pd.pivot_table(df, values='count', index=['year'],
+                    columns=['severity'], aggfunc=np.sum)
+df.columns =['not harmed','Slightly', 'harsh', 'dead']
+df_idx = df.reset_index(inplace = True)
+df.fillna(value = 0, inplace = True)
+
+df['sum_injured'] = df['harsh']+df['dead']
+N = 4
+l = []
+for x in df.sum_injured:
+    l.append(np.convolve(x, np.ones((N,))/N, mode='valid')[0])
+
+ld = pd.DataFrame(l, columns =['moving_avg'] )
+df['4_yr_moving_avg'] = ld
+df = df.drop('sum_injured', axis =1)
+df
+
+plt.style.use('seaborn-white')
+
+fig, ax = plt.subplots(figsize=(8, 4))
+
+# create a color palette
+palette = ['#f9ac86','#fc999c','#cfac86']
+
+#plt.get_cmap('Set2')
+ 
+# multiple line plot
+num=0
+for column in df.iloc[:,3:6]:
+    num+=1
+    plt.plot(df.iloc[:,0], df[column], marker='', color=palette[num-1], linewidth=1, alpha=0.9, label=column)
+
+plt.xticks([2008,2009,2010,2011,2012,2013,2014,2015,2016,2017,2018,2019])
+
+ax.get_xaxis().get_major_formatter().set_useOffset(False)
+
+
+# Add legend
+plt.legend(loc='upper left')
+
+plt.box(on=None)
+
+# Add titles
+plt.title("count of harshly injured and dead has increased in the last years", loc='left', fontsize=11, fontweight=0, color='k')
+plt.xlabel("year")
+plt.ylabel("injured count")
+
+#plt.savefig(r'slide_25.4.1.png')
+
+#graph 3
+data_08_10_250020 = data_years(data_4_250020, 2008, 2010)
+data_11_13_250020 = data_years(data_4_250020, 2011, 2013)
+data_14_16_250020 = data_years(data_4_250020, 2014, 2016)
+data_17_19_250020 = data_years(data_4_250020, 2017, 2019)
+
+df1 = data_08_10_250020.groupby('injury_severity').size() 
+df2 = data_11_13_250020.groupby('injury_severity').size()
+df3 = data_14_16_250020.groupby('injury_severity').size()
+df4 = data_17_19_250020.groupby('injury_severity').size()
+df1, df2, df3, df4
+
+# Create bars
+barWidth = 0.5
+bars1 = df1[:]
+bars2 = df2[:]
+bars3 = df3[:]
+bars4 = df4[:]
+bars5 = bars1 + bars2 + bars3 +bars4
+ 
+# The X position of bars
+r1 = [1,1.5]
+r2 = [3,3.5]
+r3 = [5,5.5]
+r4 = [7,7.5]
+r5 = r1 + r2 + r3 +r4
+
+#figure size
+plt.figure(figsize=(8,4))
+
+# Create barplot
+plt.bar(r1[0], bars1[2], width = barWidth, color = ('#f9f99f'))
+plt.bar(r1[1], bars1[3], width = barWidth, color = ('#f9ac8f'))
+plt.bar(r2[0], bars2[2], width = barWidth, color = ('#f9f99f'))
+plt.bar(r2[1], bars2[3], width = barWidth, color = ('#f9ac8f'))
+plt.bar(r3[0], bars3[2], width = barWidth, color = ('#f9f99f'))
+plt.bar(r3[1], bars3[3], width = barWidth, color = ('#f9ac8f'))
+plt.bar(r4[0], bars4[2], width = barWidth, color = ('#fbf981'), label='harsh')
+plt.bar(r4[1], bars4[3], width = barWidth, color = ('#fbac71'), label='deadly')
+ 
+# Create legend
+plt.legend(loc='upper left')
+
+# Text below each barplot
+plt.xticks([r + barWidth*5 for r in [-1.2, 0.8, 2.8, 4.8]], [' 2008-10 ', ' 2011-13 ', '  2014-16 ', '2017-19'])
+ 
+# Create labels
+label = ['11', '5', '11', '5', '11', '8', '14', '9']
+ 
+# Text on the top of each barplot
+for i in range(len(r5)):
+    plt.text(x = r5[i]-0.1 , y = 3, s = label[i], size = 11)
+
+#Adjust the margins
+plt.subplots_adjust(bottom= 0.5, top = 0.95)
+ 
+plt.title(" On road 25 harsh and deadly injured increased significantly over years", loc='left', fontsize=12, fontweight=0, color=('k'))
+
+plt.box(on=None)
+
+plt.savefig(r'slide_25.4.2.png')
+
+# Show graphic
+plt.show()
+
+#graph 12
+'''
+significant differencess were found for accident of collision with a still object
+the folowing graphs are presenting the differences (in percents) between road 250020 to Inter-urban roads count of accidents
+of collision with a still object.
+collisions with a still object are twice more common on road 250020
+'''
+data_filtered_4_250020['collision_still'] = data_filtered_4_250020['accident_type'].replace(to_replace = {1:1,2:1,3:1,4:1,5:1,7:1,8:2,9:1,10:1,11:1,12:1,13:1,15:1,17:1,19:1})
+df1 = data_filtered_4_250020.groupby(['collision_still']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+
+data_filtered_4['collision_still'] = data_filtered_4['accident_type'].replace(to_replace = {1:1,2:1,3:1,4:1,5:1,6:1,7:1,8:2,9:1,10:1,11:1,12:1,13:1,14:1,15:1,16:1,17:1,18:1,19:1,20:1})
+df2 = data_filtered_4.groupby(['collision_still']).size()
+table = pd.DataFrame(df2)
+df_idx = table.reset_index(inplace = False)
+df.insert(1,'4',table)
+
+df
+
+# Data to plot
+labels = 'collision_still_4', 'other_4'
+sizes = [df.iloc[1,1], df.iloc[0,1]]
+colors = ['#f9ac86','#cfac86']
+explode = (0.2, 0)  # explode 1st slice
+
+# Plot
+plt.pie(sizes, explode=explode, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+#plt.savefig(r'slide_25.5.1.png')
+
+plt.show()
+
+# Data to plot
+labels = 'collision_still_250020', 'other_250020'
+sizes = [df.iloc[1,0], df.iloc[0,0]]
+colors = ['#f9ac86','#cfac86']
+explode = (0.2, 0)  # explode 1st slice
+
+# Plot
+plt.pie(sizes, explode=explode, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+#plt.savefig(r'slide25.5.2.png')
+
+plt.show()
+
+#graph 12
+df1 = data_filtered_4_250020.groupby(['road_object']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+# Data to plot
+labels = 'pole', 'road sign', 'buliding', 'safety fence', 'other'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0], df.iloc[4,0]]
+colors = ['#fc999f','#f9fc99','#f9ac86','#cfac86', '#afac86']
+
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident collision objects road 25", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.5.3.png')
+
+plt.show()
+
+df1 = data_filtered_4.groupby(['road_object']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+# Data to plot
+labels = 'pole', 'road sign', 'buliding', 'safety fence', 'other'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0], df.iloc[4,0]]
+colors = ['#fc999f','#f9fc99','#f9ac86','#cfac86', '#afac86']
+
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident collision objects all interurban", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.5.4.png')
+
+plt.show()
+
+df1 =data_filtered_4_250020.groupby(['object_distance']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['count_road_25']
+df2 = data_filtered_4.groupby(['object_distance', 'object_distance_hebrew']).size()
+table = pd.DataFrame(df2)
+df_idx = table.reset_index(inplace = True)
+df.insert(1,'count_other_interurban',table[0])
+df.insert(0,'object_distance',table.object_distance_hebrew)
+df
+
+#creating data of all roads including 25
+data_4_all = road_type(data, 4)
+data_filtered_4_all = road_type(data_filtered, 4)
+
+lane1 = (data_filtered_4_all['one_lane'] > 1) & (data_filtered_4_all['one_lane'] < 4)
+df1 = data_filtered_4_all[lane1].groupby(['accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_count']
+# Data to plot
+labels =  'minor','harsh','fatal'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0]]
+colors = ['#bfa9a6','#cc99af','#f9a1a6']
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident severity on one lane roads - all interurban", loc='left', fontsize=10.5, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.6.1.png')
+plt.show()
+
+lane2 = (data_filtered_4_all['multi_lane'] > 1) & (data_filtered_4_all['multi_lane'] < 5)
+df1 = data_filtered_4_all[lane2].groupby(['accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_count']
+# Data to plot
+labels =  'minor','harsh','fatal'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0]]
+colors = ['#bfa9a6','#cc99af','#f9a1a6']
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("accident severity on multi lane roads - all interurban", loc='left', fontsize=10.5, fontweight=0, color='k')
+#plt.savefig(r'slide_25.6.2.png')
+plt.show()
+
+data_filtered_4_all.groupby(['road_object','accident_severity']).size()
+
+
+data_filtered_4_all['safety_fence'] = data_filtered_4_all['road_object'].replace(to_replace = {1:1,2:1,3:1,4:1,5:1,6:2,7:1,8:1,9:3})
+df1 = data_filtered_4_all.groupby(['safety_fence', 'accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['accident_type','accident_severity','accident_count']
+
+df
+
+# Data
+r = [0,1,2]
+df1 = df.iloc[[0,1,2],2]
+df1 = df1.reset_index(drop = True)
+df2 = df.iloc[[3,4,5],2]
+df2 = df2.reset_index(drop = True)
+df3 = df.iloc[[6,7,8],2]
+df3 = df3.reset_index(drop = True)
+df1 = pd.DataFrame(df1)
+df1.insert(1,'fatal',df3)
+df1.insert(1,'harsh',df2)
+raw_data = {'firstBars':df1.iloc[:,2],'secondBars': df1.iloc[:,0], 'thirdBars':df1.iloc[:,1]}
+df4 = pd.DataFrame(raw_data)
+
+# From raw value to percentage
+totals = [i+j+k for i,j,k in zip(df4['firstBars'], df4['secondBars'], df4['thirdBars'])]
+firstBars = [i / j * 100 for i,j in zip(df4['firstBars'], totals)]
+secondBars = [i / j * 100 for i,j in zip(df4['secondBars'], totals)]
+thirdBars = [i / j * 100 for i,j in zip(df4['thirdBars'], totals)]
+ 
+# plot
+barWidth = 0.85
+names = ('other_accidents','collision_other','collision_safety_fence')
+# Create first Bars
+plt.bar(r, firstBars, color='#bfa9a6', edgecolor='white', width=barWidth, label="minor")
+# Create second Bars
+plt.bar(r, secondBars, bottom=firstBars, color='#cc99af', edgecolor='white', width=barWidth, label="harsh")
+# Create third Bars
+plt.bar(r, secondBars, bottom=[i+j for i,j in zip(firstBars, secondBars)], color='#f9a1a6', edgecolor='white', width=barWidth, label="fatal")
+
+# Custom x axis
+plt.xticks(r, names)
+plt.xlabel("accident type")
+
+# Text of each barplot
+label1 = ['95%','87%','90%']
+label2 = ['4%', '10%', '8%']
+label3 = ['1%', '3%', '2%']
+
+
+for i in range(len(r)):
+    plt.text(x = r[i]-0.1 , y = 60, s = label1[i], size = 10)
+    plt.text(x = r[i]-0.1 , y = 89, s = label2[i], size = 10)
+    plt.text(x = r[i]-0.1 , y =98, s = label3[i], size = 10)
+
+# Add a legend
+plt.legend(loc='upper left', bbox_to_anchor=(1,1), ncol=1)
+
+plt.title("90% percent of the collision with safety fence accidents are minor \n all inter urban roads", loc='left', fontsize=12, fontweight=0, color=('k'))
+
+plt.box(on=None)
+plt.yticks([])
+
+#plt.savefig(r'slide_25.7.1.png', bbox_inches='tight')
+# Show graphic
+plt.show()
+
+#data_4_all = road_type(data, 4)
+#data_filtered_4_all = road_type(data_filtered, 4)
+
+df1 = data_filtered_4_all.groupby(['accident_type','accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['accident_type','accident_severity','accident_count']
+df = pd.pivot_table(df, values='accident_count', index=['accident_type'],
+                    columns=['accident_severity'], aggfunc=np.sum)
+df.columns =['minor','harsh', 'fatal']
+
+a = []
+for i in range(19):
+    b = df.iloc[i,:].sum()
+    for j in range(3):
+        c = df.iloc[i,j]/b
+        a.append(c.round(2))
+ar = np.array(a)
+ar = ar.reshape(19, 3)
+
+df1 = pd.DataFrame(ar)
+df1.index = df.index
+df1.columns = ['minor', 'harsh', 'fatal']
+
+df1 = data_filtered_4_all.groupby(['accident_type','accident_severity']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['accident_type','accident_severity','accident_count']
+df = pd.pivot_table(df, values='accident_count', index=['accident_type'],
+                    columns=['accident_severity'], aggfunc=np.sum)
+df.columns =['minor','harsh', 'fatal']
+df = df.fillna(value = 0)
+
+# Data to plot
+labels = ['pedestrian', 'front side collision', 'Front to back', 'Side-side collision','Frontal Collision', 'Collision stopped vehicle','Collision parked vehicle', 'still collision', 'down road or sidewalk', 'overturning', 'skided', 'Injury to a passenger inside the vehicle', 'fall moving vehicle', 'back to front', 'back to side', 'animal', "car's luggage"]
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0], df.iloc[4,0], df.iloc[5,0], df.iloc[6,0], df.iloc[7,0], df.iloc[8,0], df.iloc[9,0],df.iloc[10,0], df.iloc[11,0], df.iloc[12,0], df.iloc[15,0], df.iloc[16,0], df.iloc[17,0], df.iloc[18,0]]
+colors = ['#9eab99','#f9ac86','#fc999f','#f9fc99','#afac86','#e9acd9','#cfac86','#bc999f','#b9fc99','#b9eff9','#a294f9','#fcac86','#fcc99f','#dc2c99','#bc1c86','#eeabb6','#9e2bb3']
+
+# Plot
+patches, texts, pct = plt.pie(sizes, colors=colors, autopct='%1d%%', startangle=90)
+plt.legend(patches, labels, loc=(0.85, 0.12))
+
+plt.axis('equal')
+
+plt.title("minor accident types", loc='left', fontsize=13, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.11.3.png')
+
+plt.show()
+
+# Data to plot
+labels = ['pedestrian', 'front side collision', 'Front to back', 'Side-side collision','Frontal Collision', 'Collision stopped vehicle','Collision parked vehicle', 'still collision', 'down road or sidewalk', 'overturning', 'skided', 'Injury to a passenger inside the vehicle', 'fall moving vehicle', 'back to front', 'back to side', 'animal', "car's luggage"]
+sizes = [df.iloc[0,1], df.iloc[1,1], df.iloc[2,1], df.iloc[3,1], df.iloc[4,1], df.iloc[5,1], df.iloc[6,1], df.iloc[7,1], df.iloc[8,1], df.iloc[9,1],df.iloc[10,1], df.iloc[11,1], df.iloc[12,1], df.iloc[15,1], df.iloc[16,1], df.iloc[17,1], df.iloc[18,1]]
+colors = ['#9eab99','#f9ac86','#fc999f','#f9fc99','#afac86','#e9acd9','#cfac86','#bc999f','#b9fc99','#b9eff9','#a294f9','#fcac86','#fcc99f','#dc2c99','#bc1c86','#eeabb6','#9e2bb3']
+
+# Plot
+patches, texts, pct = plt.pie(sizes, colors=colors, autopct='%1d%%', startangle=90)
+plt.legend(patches, labels, loc=(0.85, 0.12))
+
+plt.axis('equal')
+
+plt.title("harsh accident types", loc='left', fontsize=13, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.11.3.png')
+
+plt.show()
+
+# Data to plot
+labels = ['pedestrian', 'front side collision', 'Front to back', 'Side-side collision','Frontal Collision', 'Collision stopped vehicle','Collision parked vehicle', 'still collision', 'down road or sidewalk', 'overturning', 'skided', 'Injury to a passenger inside the vehicle', 'fall moving vehicle', 'back to front', 'back to side', 'animal', "car's luggage"]
+sizes = [df.iloc[0,2], df.iloc[1,2], df.iloc[2,2], df.iloc[3,2], df.iloc[4,2], df.iloc[5,2], df.iloc[6,2], df.iloc[7,2], df.iloc[8,2], df.iloc[9,2],df.iloc[10,2], df.iloc[11,2], df.iloc[12,2], df.iloc[15,2], df.iloc[16,2], df.iloc[17,2], df.iloc[18,2]]
+colors = ['#9eab99','#f9ac86','#fc999f','#f9fc99','#afac86','#e9acd9','#cfac86','#bc999f','#b9fc99','#b9eff9','#a294f9','#fcac86','#fcc99f','#dc2c99','#bc1c86','#eeabb6','#9e2bb3']
+
+# Plot
+patches, texts, pct = plt.pie(sizes, colors=colors, autopct='%1d%%', startangle=90)
+plt.legend(patches, labels, loc=(0.85, 0.12))
+
+plt.axis('equal')
+
+plt.title("fatal accident types", loc='left', fontsize=13, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.11.3.png')
+
+plt.show()
+
+'''
+significant differencess were found for front to side collision. 
+the folowing graph is comparing the precents of accident front side collisions and other types of accidents 
+by accidents types on road 250020.
+accident collisions relative part of harsh and fatal accidents is bigger than minor ones.
+'''
+#data_filtered_4_250020['front_side_collision'] = data_filtered_4_250020['accident_type'].replace(to_replace = {1:1,2:2,3:1,4:1,5:1,7:1,8:1,9:1,10:1,11:1,12:1,13:1,15:1,17:1,19:1})
+df1 = data_filtered_4_250020.groupby(['accident_severity','front_side_collision']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['accident_severity','front_side_collision','count']
+
+df
+
+
+from matplotlib import rc
+# Data
+r = [0,1,2]
+df1 = df.iloc[[1,3,5],2]
+df1 = df1.reset_index(drop = True)
+df2 = df.iloc[[0,2,4],2]
+df2 = df2.reset_index(drop = True)
+df1 = pd.DataFrame(df1)
+df1.insert(1,'count all',df2)
+raw_data = {'firstBars': df1.iloc[:,0], 'secondBars':df1.iloc[:,1]}
+df3 = pd.DataFrame(raw_data)
+ 
+# From raw value to percentage
+totals = [i+j for i,j in zip(df3['firstBars'], df3['secondBars'])]
+firstBars = [i / j * 100 for i,j in zip(df3['firstBars'], totals)]
+secondBars = [i / j * 100 for i,j in zip(df3['secondBars'], totals)]
+
+# plot
+barWidth = 0.85
+names = ('minor','harsh','fatal')
+# Create green Bars
+plt.bar(r, firstBars, color='#f3bc87', edgecolor='white', width=barWidth, label="front-side collision")
+# Create orange Bars
+plt.bar(r, secondBars, bottom=firstBars, color='#f6ffb5', edgecolor='white', width=barWidth, label="other")
+
+# Custom x axis
+plt.xticks(r, names)
+plt.xlabel("accident severity")
+
+# Text of each barplot
+label1 = ['22%','24%','52%']
+label2 = ['78%', '76%', '48%']
+
+for i in range(len(r)):
+    plt.text(x = r[i]-0.1 , y = 9, s = label1[i], size = 10)
+    plt.text(x = r[i]-0.1 , y = 60, s = label2[i], size = 10)
+
+# Add a legend
+plt.legend(loc='upper left', bbox_to_anchor=(1,1), ncol=1)
+
+plt.title("more than 50% of the fatal accidents are front-side collision on road 250020", loc='left', fontsize=12, fontweight=0, color=('k'))
+
+plt.box(on=None)
+plt.yticks([])
+
+#plt.savefig(r'slide_25.7.2.png', bbox_inches='tight')
+# Show graphic
+plt.show()
+
+#data_filtered_4_250020['front_side_collision'] = data_filtered_4_250020['accident_type'].replace(to_replace = {1:1,2:2,3:1,4:1,5:1,7:1,8:1,9:1,10:1,11:1,12:1,13:1,15:1,17:1,19:1})
+df1 = data_filtered_4_250020.groupby(['day_night','front_side_collision']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = True)
+df.columns =['day_night','front_side_collision','count']
+df = pd.pivot_table(df, values='count', index=['day_night'],
+                    columns=['front_side_collision'], aggfunc=np.sum)
+df.columns =['Other_250020','front_side_collision_250020']
+#df = df.apply(lambda x: x/x.sum()).round(2)
+df
+
+data_filtered_4['front_side_collision'] = data_filtered_4['accident_type'].replace(to_replace = {1:1,2:2,3:1,4:1,5:1,6:1,7:1,8:1,9:1,10:1,11:1,12:1,13:1,14:1,15:1,16:1,17:1,18:1,19:1,20:1})
+df2 = data_filtered_4.groupby(['day_night','front_side_collision']).size()
+df3 = pd.DataFrame(df2)
+df_idx = df3.reset_index(inplace = True)
+df3.columns =['day_night','front_side_collision','count']
+df3 = pd.pivot_table(df3, values='count', index=['day_night'],
+                    columns=['front_side_collision'], aggfunc=np.sum)
+df3.columns =['Other_4','front_side_collision_4']
+#df3 = df3.apply(lambda x: x/x.sum()).round(2)
+df3
+
+# Data
+r = [0,1]
+raw_data = {'firstBars': df.front_side_collision_250020, 'secondBars':df3.front_side_collision_4}
+df4 = pd.DataFrame(raw_data)
+# From raw value to percentage
+totals = [i+j for i,j in zip(df4['firstBars'], df4['secondBars'])]
+firstBars = [i / j * 100 for i,j in zip(df4.firstBars, totals)]
+secondBars = [i / j * 100 for i,j in zip(df4.secondBars, totals)]
+ 
+# plot
+barWidth = 0.85
+names = ('road 25','inter urban roads')
+# Create first Bars
+plt.bar(r, firstBars, color='khaki', edgecolor='white', width=barWidth, label="day")
+# Create second Bars
+plt.bar(r, secondBars, bottom=firstBars, color='dimgray', edgecolor='white', width=barWidth, label="night")
+
+# Custom x axis
+plt.xticks(r, names)
+plt.xlabel("road")
+
+# Text of each barplot
+label1 = ['63%','70%']
+label2 = ['37%', '30%']
+
+for i in range(len(r)):
+    plt.text(x = r[i]-0.1 , y = 25, s = label1[i], size = 10)
+    plt.text(x = r[i]-0.1 , y = 60, s = label2[i], size = 10)
+
+# Add a legend
+plt.legend(loc='upper left', bbox_to_anchor=(1,1), ncol=1)
+
+plt.title("on road 250020 front side collision accidents accurr more often at night", loc='left', fontsize=12, fontweight=0, color=('k'))
+
+plt.box(on=None)
+plt.yticks([])
+
+#plt.savefig(r'slide_25.8.png', bbox_inches='tight')
+# Show graphic
+plt.show()
+
+#map of front side accidents, more common near juctions
+accu_frontside = (data_filtered_4_250020['accident_type'] == 2) & (data_filtered_4_250020['location_accuracy'] == 1)
+data_filtered_4_250020[accu_frontside].groupby(['latitude','longitude'])
+
+import folium
+m = folium.Map(location=[31.198543, 34.923482], zoom_start=12)
+for row in data_filtered_4_250020[accu_frontside][['longitude','latitude']].values:
+    try:
+        folium.Circle(radius=10,
+                      location = [row[1], row[0]]
+        ).add_to(m)
+    except ValueError:
+        continue
+
+display(m)
+
+
+'''
+significant differencess were found for overturn accidents
+the folowing graphs are presenting the differences (in percents) between road 250020 to Inter-urban roads count of overturn accidents.
+overturns are more common on road 250020
+'''
+data_filtered_4_250020['overturned'] = data_filtered_4_250020['accident_type'].replace(to_replace = {1:1,2:1,3:1,4:1,5:1,7:1,8:1,9:1,10:2,11:1,12:1,13:1,15:1,17:1,19:1})
+df1 = data_filtered_4_250020.groupby(['overturned']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+
+data_filtered_4['overturned'] = data_filtered_4['accident_type'].replace(to_replace = {1:1,2:1,3:1,4:1,5:1,6:1,7:1,8:1,9:1,10:2,11:1,12:1,13:1,14:1,15:1,16:1,17:1,18:1,19:1,20:1})
+df2 = data_filtered_4.groupby(['overturned']).size()
+table = pd.DataFrame(df2)
+df_idx = table.reset_index(inplace = False)
+df.insert(1,'4',table)
+
+df
+
+# prepare data
+names='overturned_250020', 'other_250020'
+size=[df.iloc[1,0],df.iloc[0,0].sum()]
+ 
+# Create a circle for the center of the plot
+my_circle=plt.Circle((0,0), 0.7, color='white')
+
+# Give color names
+plt.pie(size, labels=names,autopct='%1.1f%%', wedgeprops = { 'linewidth' : 0.5, 'edgecolor' : 'white' }, colors=['#fb999f','#f6ffb5'])
+p=plt.gcf()
+p.gca().add_artist(my_circle)
+
+#plt.savefig(r'slide_25.10.1.png', bbox_inches='tight')
+
+# prepare data
+names='overturned_4', 'other_4'
+size=[df.iloc[1,1],df.iloc[0,1].sum()]
+ 
+# Create a circle for the center of the plot
+my_circle=plt.Circle((0,0), 0.7, color='white')
+
+# Give color names
+plt.pie(size, labels=names,autopct='%1.1f%%', wedgeprops = { 'linewidth' : 0.5, 'edgecolor' : 'white' }, colors=['#fb999f','#f6ffb5'])
+p=plt.gcf()
+p.gca().add_artist(my_circle)
+
+#plt.savefig(r'slide_25.10.2.png', bbox_inches='tight')
+
+#7
+'''
+
+'''
+truck = data_4_250020['accident_type'] == 10
+data_4_250020['involve_vehicle_type2'] = data_4_250020['involve_vehicle_type'].replace(to_replace = {1:1,2:2,3:3,4:3,5:1,6:3})
+df1 = data_4_250020[truck].groupby(['involve_vehicle_type2']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_type_250020']
+
+truck4 = data_4['accident_type'] == 10
+data_4['involve_vehicle_type2'] = data_4['involve_vehicle_type'].replace(to_replace = {1:1,2:2,3:3,4:3,5:3,6:3,7:3,8:3,9:3,10:3})
+df2 = data_4[truck4].groupby(['involve_vehicle_type2']).size()
+table = pd.DataFrame(df2)
+df_idx = table.reset_index(inplace = False)
+df.insert(1,'accident_type_4',table)
+
+#df = df.apply(lambda x: x/x.sum())
+
+df
+
+#graph 7
+ 
+# Create bars
+barWidth = 0.9
+bars1 = df.iloc[0,:]
+bars2 = df.iloc[1,:]
+bars3 = df.iloc[2,:]
+bars4 = df
+ 
+# The X position of bars
+r1 = [1,2]
+r2 = [3,4]
+r3 = [5,6]
+r4 = r1 + r2 + r3  
+ 
+                                                  
+# Create barplot
+plt.bar(r1[0], bars1[0], width = barWidth, color = ('#f9f99f'))
+plt.bar(r1[1], bars1[1], width = barWidth, color = ('#f9ac8f'))
+plt.bar(r2[0], bars2[0], width = barWidth, color = ('#fbf970'), label='250020')
+plt.bar(r2[1], bars2[1], width = barWidth, color = ('#fbac60'), label='4')
+plt.bar(r3[0], bars3[0], width = barWidth, color = ('#f9f99f'))
+plt.bar(r3[1], bars3[1], width = barWidth, color = ('#f9ac8f'))
+# Note: the barplot could be created easily. See the barplot section for other examples.
+ 
+# Create legend
+plt.legend(loc = 'upper right')
+
+# Text below each barplot with a rotation at 90°
+plt.xticks([r + barWidth for r in range((6))], ['Private', 'Private', 'Truck', 'Truck', 'Other', 'Other'], rotation=90)
+ 
+# Create labels
+label = ['61%', '62%', '30%', '21%', '8%', '16%']
+y = [0.61, 0.62, 0.30, 0.21, 0.08, 0.16]
+
+# Text on the top of each barplot
+for i in range(len(r4)):
+    plt.text(x = r4[i]-0.3 , y = y[i]-0.03, s = label[i], size = 10, color = 'k')
+
+#Adjust the margins
+plt.subplots_adjust(bottom= 0.2, top = 0.98)
+
+#title
+plt.title("Trucks involved in overturned accidents on road 250020 more than on other inter urban roads", loc='left', fontsize=11, fontweight=0, color='k')
+
+plt.box(on=None)
+plt.yticks([])
+
+#plt.savefig(r'slide_25.11.1.png', bbox_inches='tight')
+
+# Show graphic
+plt.show()
+
+overturn = data_filtered_4_250020['accident_type'] == 10
+df1 = data_filtered_4_250020[overturn].groupby(['involve_vehicle_type_t']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_type_250020']
+# Data to plot
+labels = 'private', 'Cargo up to 3.5 tons - Unified', 'Cargo up to 3.5 tons - not unified', 'Freight 3.6 to 34 tons','other'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0], df.iloc[4,0] ]
+colors = ['#cfac86','#f9ac86','#fc999f','#f9fc99','#afac86']
+
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("Trucks types involved in overturned accidents on road 250020\n", loc='left', fontsize=11, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.11.2.png')
+
+plt.show()
+
+notoverturn = data_filtered_4_250020['accident_type'] != 10
+df1 = data_filtered_4_250020[notoverturn].groupby(['involve_vehicle_type_t']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_type_250020']
+# Data to plot
+labels = 'private', 'Cargo up to 3.5 tons - Unified', 'Cargo up to 3.5 tons - not unified', 'Freight 3.6 to 34 tons','other'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0], df.iloc[4,0] ]
+colors = ['#cfac86','#f9ac86','#fc999f','#f9fc99','#afac86']
+
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+plt.title("Trucks types involved in overturned accidents on other urban inter roads\n", loc='left', fontsize=11, fontweight=0, color='k')
+
+#plt.savefig(r'slide_25.11.3.png')
+
+plt.show()
+
+animal = data_filtered_4_250020['accident_type'] == 19
+df1 = data_filtered_4_250020[animal].groupby(['day_night']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['accident_type_250020']
+
+animal4 = data_filtered_4['accident_type'] == 19
+df2 = data_filtered_4[animal4].groupby(['day_night']).size()
+table = pd.DataFrame(df2)
+df_idx = table.reset_index(inplace = False)
+df.insert(1,'accident_type_4',table)
+
+df = df.apply(lambda x: x/x.sum())
+
+df
+
+# Data
+r = [0,1]
+raw_data = {'firstBars': df.iloc[0,:], 'secondBars':df.iloc[1,:]}
+df4 = pd.DataFrame(raw_data)
+ 
+# From raw value to percentage
+totals = [i+j for i,j in zip(df4['firstBars'], df4['secondBars'])]
+firstBars = [i / j * 100 for i,j in zip(df4['firstBars'], totals)]
+secondBars = [i / j * 100 for i,j in zip(df4['secondBars'], totals)]
+ 
+# plot
+barWidth = 0.85
+names = ('250020','inter urban roads')
+# Create green Bars
+plt.bar(r, firstBars, color='khaki', edgecolor='white', width=barWidth, label="day")
+# Create orange Bars
+plt.bar(r, secondBars, bottom=firstBars, color='dimgray', edgecolor='white', width=barWidth, label="night")
+
+# Custom x axis
+plt.xticks(r, names)
+plt.xlabel("road")
+
+# Text of each barplot
+label1 = ['50%','16%']
+label2 = ['50%', '84%']
+
+for i in range(len(r)):
+    plt.text(x = r[i]-0.1 , y = 10, s = label1[i], size = 10)
+    plt.text(x = r[i]-0.1 , y = 75, s = label2[i], size = 10)
+
+# Add a legend
+plt.legend(loc='upper left', bbox_to_anchor=(1,1), ncol=1)
+
+plt.title("on road 25 collision with an animal accurr more often at day", loc='left', fontsize=12, fontweight=0, color=('k'))
+
+plt.box(on=None)
+plt.yticks([])
+
+#plt.savefig(r'slide_25.12.png', bbox_inches='tight')
+
+# Show graphic
+plt.show()
+
+pedestrian = data_250020['accident_type'] == 1
+df1 = data_250020[pedestrian].groupby(['population_type']).size()
+df = pd.DataFrame(df1)
+df_idx = df.reset_index(inplace = False)
+df.columns =['250020']
+# Data to plot
+labels = 'Jewish', 'Arab', 'Other', 'Forighn'
+sizes = [df.iloc[0,0], df.iloc[1,0], df.iloc[2,0], df.iloc[3,0]]
+colors = ['#f9ac86','#fc999f','#cfac86','#f9fc99']
+#explode = (0.2, 0)  # explode 1st slice
+
+# Plot
+plt.pie(sizes, labels=labels, colors=colors,
+autopct='%1d%%', shadow=True, startangle=140)
+
+plt.axis('equal')
+
+#plt.savefig(r'slide25.13.1.png', bbox_inches='tight')
+
+plt.show()
+
+#most of the pedestrian hits happened near beduian settelments junctions
+
+pedestrian = data_250020['accident_type'] == 1
+data_250020[pedestrian].groupby(['latitude','longitude'])
+
+import folium
+m = folium.Map(location=[31.198543, 34.923482], zoom_start=11)
+for row in data_250020[pedestrian][['longitude','latitude']].values:
+    try:
+        folium.Circle(radius=10,
+                      location = [row[1], row[0]]
+        ).add_to(m)
+    except ValueError:
+        continue
+
+display(m)
+
+
+
+