Happiness Index (Introduction)¶
World Happiness Index Report¶
Context
The World Happiness Report is a universal survey on the state of global happiness. First published in 2012, it was updated until 2016. Since then, it has been used in various fields: economics, psychology, national statistics, healthcare, universal insurance, and others.
Happiness scores use data from the Gallup World Poll. The scores are based on evaluations of main life issues assessed through survey questions, known as the Cantril ladder.
The happiness score includes the following six factors determining survey results:
- Economy
- Social support
- Life expectancy
- Freedom
- Absence of corruption
- Well-being
These factors determine how happy people are in each country compared to people living in "dystopia.
Inspiration
- Which countries or regions rank highest in overall happiness?
- Do these 6 factors affect happiness?
- Are the components of happiness correlated with suicide cases?
- Is it possible to predict the number of suicide cases if we know the components of the happiness index?
And other questions that will arise later.
What is Dystopia?
Dystopia is an imaginary country where the unhappiest people in the world live. Each surveyed person compares their situation to such a country, where there is the lowest income, minimal life expectancy, low well-being, high corruption, lack of freedom, and minimal social support. The comparison is made with the opposite of utopia, "dystopia."
Which columns describe the Happiness Index?
The following columns:
- GDP per Capita
- Family
- Life Expectancy
- Freedom
- Generosity
- Trust Government Corruption
Dystopia Residual is Dystopia Happiness Score (1.85) + Residual.
If you sum these factors, you will get the Happiness Score.
How the data looks
In [2]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
In [3]:
# Read the 2015 data
y_2015 = pd.read_csv('data/2015.csv')
In [7]:
# Let's look at the first five records
y_2015[0:10].style.set_properties(**{'background-color': 'red'}, subset=['Happiness Score'])
Out[7]:
| Country | Region | Happiness Rank | Happiness Score | Standard Error | Economy (GDP per Capita) | Family | Health (Life Expectancy) | Freedom | Trust (Government Corruption) | Generosity | Dystopia Residual | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | Switzerland | Western Europe | 1 | 7.587000 | 0.034110 | 1.396510 | 1.349510 | 0.941430 | 0.665570 | 0.419780 | 0.296780 | 2.517380 |
| 1 | Iceland | Western Europe | 2 | 7.561000 | 0.048840 | 1.302320 | 1.402230 | 0.947840 | 0.628770 | 0.141450 | 0.436300 | 2.702010 |
| 2 | Denmark | Western Europe | 3 | 7.527000 | 0.033280 | 1.325480 | 1.360580 | 0.874640 | 0.649380 | 0.483570 | 0.341390 | 2.492040 |
| 3 | Norway | Western Europe | 4 | 7.522000 | 0.038800 | 1.459000 | 1.330950 | 0.885210 | 0.669730 | 0.365030 | 0.346990 | 2.465310 |
| 4 | Canada | North America | 5 | 7.427000 | 0.035530 | 1.326290 | 1.322610 | 0.905630 | 0.632970 | 0.329570 | 0.458110 | 2.451760 |
| 5 | Finland | Western Europe | 6 | 7.406000 | 0.031400 | 1.290250 | 1.318260 | 0.889110 | 0.641690 | 0.413720 | 0.233510 | 2.619550 |
| 6 | Netherlands | Western Europe | 7 | 7.378000 | 0.027990 | 1.329440 | 1.280170 | 0.892840 | 0.615760 | 0.318140 | 0.476100 | 2.465700 |
| 7 | Sweden | Western Europe | 8 | 7.364000 | 0.031570 | 1.331710 | 1.289070 | 0.910870 | 0.659800 | 0.438440 | 0.362620 | 2.371190 |
| 8 | New Zealand | Australia and New Zealand | 9 | 7.286000 | 0.033710 | 1.250180 | 1.319670 | 0.908370 | 0.639380 | 0.429220 | 0.475010 | 2.264250 |
| 9 | Australia | Australia and New Zealand | 10 | 7.284000 | 0.040830 | 1.333580 | 1.309230 | 0.931560 | 0.651240 | 0.356370 | 0.435620 | 2.266460 |
Data Exploration¶
In [4]:
# Use seaborn styles:
sns.set_palette("GnBu_d")
sns.set_style('whitegrid')
The impact of freedom on the happiness index:
In [5]:
sns.jointplot(x='Happiness Score',y='Freedom',data=y_2015);
The impact of economic conditions on happiness:
In [6]:
sns.relplot(x='Happiness Score',y='Economy (GDP per Capita)',data=y_2015,
hue='Region');
Let's build graphs of the interdependence of all parameters
In [68]:
# Let's build graphs of the interdependence of all parameters:
sns.pairplot(y_2015[['Happiness Score',
'Economy (GDP per Capita)', 'Family',
'Health (Life Expectancy)', 'Freedom', 'Trust (Government Corruption)',
'Generosity', 'Dystopia Residual']]);
Create a linear model plot using lmplot to show the relationship between health and happiness
In [8]:
sns.lmplot(x='Happiness Score',y='Health (Life Expectancy)',data=y_2015);
Let's see the normal distribution of the Happiness Index:
In [9]:
sns.histplot(y_2015['Happiness Score'],bins=25,kde=True);
Happiness Index by Region
In [60]:
ax = sns.barplot(y="Region", x="Happiness Score", data=y_2015)
Suicide Cases¶
In [10]:
# Import data on suicide cases by country and year:
suicide = pd.read_csv('data/share-deaths-suicide.csv')
In [11]:
# Rename the columns:
suicide=suicide[['Entity','Year','Deaths - Self-harm - Sex: Both - Age: All Ages (Percent)']].rename(columns={'Entity':'Country','Deaths - Self-harm - Sex: Both - Age: All Ages (Percent)':'suicide'})
In [12]:
# Keep only the data for the year 2015 and check the info
suicide_2015 = suicide[suicide['Year'] == 2015]
suicide_2015.head()
Out[12]:
| Country | Year | suicide | |
|---|---|---|---|
| 25 | Afghanistan | 2015 | 0.666678 |
| 53 | Albania | 2015 | 0.786902 |
| 81 | Algeria | 2015 | 1.048141 |
| 109 | American Samoa | 2015 | 1.086531 |
| 137 | Andean Latin America | 2015 | 1.128180 |
Merge with our data
In [13]:
# It seems there are no NaNs in the data, so we can merge the data:
df = pd.merge(y_2015, suicide_2015, on=['Country'])
In [14]:
# Drop the 'Year' column
df.drop(columns=['Year'], inplace=True)
df.head()
Out[14]:
| Country | Region | Happiness Rank | Happiness Score | Standard Error | Economy (GDP per Capita) | Family | Health (Life Expectancy) | Freedom | Trust (Government Corruption) | Generosity | Dystopia Residual | suicide | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | Switzerland | Western Europe | 1 | 7.587 | 0.03411 | 1.39651 | 1.34951 | 0.94143 | 0.66557 | 0.41978 | 0.29678 | 2.51738 | 1.827820 |
| 1 | Iceland | Western Europe | 2 | 7.561 | 0.04884 | 1.30232 | 1.40223 | 0.94784 | 0.62877 | 0.14145 | 0.43630 | 2.70201 | 1.884182 |
| 2 | Denmark | Western Europe | 3 | 7.527 | 0.03328 | 1.32548 | 1.36058 | 0.87464 | 0.64938 | 0.48357 | 0.34139 | 2.49204 | 1.281190 |
| 3 | Norway | Western Europe | 4 | 7.522 | 0.03880 | 1.45900 | 1.33095 | 0.88521 | 0.66973 | 0.36503 | 0.34699 | 2.46531 | 1.429625 |
| 4 | Canada | North America | 5 | 7.427 | 0.03553 | 1.32629 | 1.32261 | 0.90563 | 0.63297 | 0.32957 | 0.45811 | 2.45176 | 1.753301 |
In [16]:
# Let's build a correlation map:
sns.heatmap(df[['Happiness Score',
'Economy (GDP per Capita)', 'Family',
'Health (Life Expectancy)', 'Freedom', 'Trust (Government Corruption)',
'Generosity', 'Dystopia Residual','suicide']].corr(), annot=True)
Out[16]:
<AxesSubplot:>
ვნახოთ უფრო ახლოდან თუ როგორ არის სუიციდიი და ბედნიერების ინდექსი ურთიერთდამოკიდებული
In [70]:
# Let's take a closer look at how suicide and the happiness index are interrelated
sns.lmplot(x='Happiness Score', y='suicide', data=df);
Let's also compare it to the economic situation
In [71]:
#ასევე შევადაროთ ეკონომიკურ მდგომარეობას
sns.lmplot(x='Economy (GDP per Capita)',y='suicide',data=df);
As seen from the graphs, there is a slight but noticeable correlation between suicide and economic conditions. This means that in economically developed countries, there were proportionally more cases of suicide during the surveyed period.
In [72]:
sns.relplot(data=df,x='Economy (GDP per Capita)',y='Happiness Score',hue='Region',size="suicide",sizes=(1, 200));
In [20]:
df.sort_values("suicide",ascending=False).head(10)
Out[20]:
| Country | Region | Happiness Rank | Happiness Score | Standard Error | Economy (GDP per Capita) | Family | Health (Life Expectancy) | Freedom | Trust (Government Corruption) | Generosity | Dystopia Residual | suicide | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 45 | South Korea | Eastern Asia | 47 | 5.984 | 0.04098 | 1.24461 | 0.95774 | 0.96538 | 0.33208 | 0.07857 | 0.18557 | 2.21978 | 5.595541 |
| 27 | Qatar | Middle East and Northern Africa | 28 | 6.611 | 0.06257 | 1.69042 | 1.07860 | 0.79733 | 0.64040 | 0.52208 | 0.32573 | 1.55674 | 3.970951 |
| 122 | Sri Lanka | Southern Asia | 132 | 4.271 | 0.03751 | 0.83524 | 1.01905 | 0.70806 | 0.53726 | 0.09179 | 0.40828 | 0.67108 | 3.597424 |
| 38 | Suriname | Latin America and Caribbean | 40 | 6.269 | 0.09811 | 0.99534 | 0.97200 | 0.60820 | 0.59657 | 0.13633 | 0.16991 | 2.79094 | 3.492400 |
| 52 | Kazakhstan | Central and Eastern Europe | 54 | 5.855 | 0.04114 | 1.12254 | 1.12241 | 0.64368 | 0.51649 | 0.08454 | 0.11827 | 2.24729 | 3.132291 |
| 54 | Lithuania | Central and Eastern Europe | 56 | 5.833 | 0.03843 | 1.14723 | 1.25745 | 0.73128 | 0.21342 | 0.01031 | 0.02641 | 2.44649 | 2.655972 |
| 23 | Singapore | Southeastern Asia | 24 | 6.798 | 0.03780 | 1.52186 | 1.02000 | 1.02525 | 0.54252 | 0.49210 | 0.31105 | 1.88501 | 2.641140 |
| 93 | Mongolia | Eastern Asia | 100 | 4.874 | 0.03313 | 0.82819 | 1.30060 | 0.60268 | 0.43626 | 0.02666 | 0.33230 | 1.34759 | 2.599839 |
| 19 | United Arab Emirates | Middle East and Northern Africa | 20 | 6.901 | 0.03729 | 1.42727 | 1.12575 | 0.80925 | 0.64157 | 0.38583 | 0.26428 | 2.24743 | 2.540437 |
| 36 | Taiwan | Eastern Asia | 38 | 6.298 | 0.03868 | 1.29098 | 1.07617 | 0.87530 | 0.39740 | 0.08129 | 0.25376 | 2.32323 | 2.499941 |
2015-2020 პერიოდის ანალიზი¶
In [21]:
y_2016 = pd.read_csv('data/2016.csv')
y_2017 = pd.read_csv('data/2017.csv')
y_2018 = pd.read_csv('data/2018.csv')
y_2019 = pd.read_csv('data/2019.csv')
y_2020 = pd.read_csv('data/2020.csv')
ვნახოთ როგორ იცვლებოდა ჯანმრთელობის ინდექსი წლების განმაცლობაში¶
In [22]:
plt.figure(figsize=(10,5))
sns.kdeplot(y_2015['Health (Life Expectancy)'],color='black')
sns.kdeplot(y_2016['Health (Life Expectancy)'],color='blue')
sns.kdeplot(y_2017['Health..Life.Expectancy.'],color='limegreen')
sns.kdeplot(y_2018['Healthy life expectancy'],color='orange')
sns.kdeplot(y_2019['Healthy life expectancy'],color='pink')
sns.kdeplot(y_2020['Explained by: Healthy life expectancy'],color='red')
plt.title('Health over the Years',size=20)
plt.legend([2015,2016,2017,2018,2019,2020])
plt.show()
ბედნიერების ინდექსის ცვლილება :¶
In [23]:
plt.figure(figsize=(10,5))
sns.kdeplot(y_2015['Happiness Score'],color='black')
sns.kdeplot(y_2016['Happiness Score'],color='blue')
sns.kdeplot(y_2017['Happiness.Score'],color='limegreen')
sns.kdeplot(y_2018['Score'],color='orange')
sns.kdeplot(y_2019['Score'],color='pink')
sns.kdeplot(y_2020['Ladder score'],color='red')
plt.title('Happiness over the Years',size=20)
plt.legend([2015,2016,2017,2018,2019,2020])
plt.show()
Geopandas-ს გამოყენებით რუკაზე მონაცემების გამოსახვა¶
In [24]:
import geopandas as gpd
sns.set_style("darkgrid")
import math
map_df = gpd.read_file('data/World_Countries__Generalized_.shp',)
map_df = map_df.replace({'Russian Federation':'Russia',
'Trinidad and Tobago': 'Trinidad & Tobago',
"Côte d'Ivoire": 'Ivory Coast',
'Congo': 'Congo (Brazzaville)',
'Congo DRC':'Congo (Kinshasa)',
'Palestinian Territory':'Palestinian Territories'})
In [25]:
data_2018 = y_2018.rename(index = str, columns = {'Country or region':"Country"})
df_2018 = data_2018[['Country','Score']]
merged = map_df.set_index('COUNTRY').join(df.set_index('Country'))
In [26]:
variable = 'Score'
vmin, vmax = 2.853,7.021
In [27]:
df_2018=df_2018.set_index('Country').T.to_dict('list')
updates = {'Maldives':[5.20],
'Oman':[6.853],
'Sudan':[4.14],
'Djibouti':[4.37],
'Angola':[3.80],
'Belize':[5.95599985122681]
}
df_2018.update(updates)
In [28]:
for i in range (len(merged)):
if math.isnan(merged['Happiness Score'][i]):
if (str(merged['COUNTRYAFF'][i])) in df_2018:
merged['Happiness Score'][i] = float(df_2018[str(merged['COUNTRYAFF'][i])][0])
merged.head(5)
<ipython-input-28-53412ebec0e1>:4: SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy merged['Happiness Score'][i] = float(df_2018[str(merged['COUNTRYAFF'][i])][0])
Out[28]:
| FID | ISO | COUNTRYAFF | AFF_ISO | SHAPE_Leng | SHAPE_Area | geometry | Region | Happiness Rank | Happiness Score | Standard Error | Economy (GDP per Capita) | Family | Health (Life Expectancy) | Freedom | Trust (Government Corruption) | Generosity | Dystopia Residual | suicide | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COUNTRY | |||||||||||||||||||
| American Samoa | 1 | AS | United States | US | 0.600124 | 0.013720 | POLYGON ((-170.74390 -14.37555, -170.74942 -14... | NaN | NaN | 6.886 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| United States Minor Outlying Islands | 2 | UM | United States | US | 0.480216 | 0.003216 | MULTIPOLYGON (((-160.02114 -0.39805, -160.0281... | NaN | NaN | 6.886 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| Cook Islands | 3 | CK | New Zealand | NZ | 0.980664 | 0.013073 | MULTIPOLYGON (((-159.74698 -21.25667, -159.793... | NaN | NaN | 7.324 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| French Polynesia | 4 | PF | France | FR | 3.930211 | 0.175332 | MULTIPOLYGON (((-149.17920 -17.87084, -149.258... | NaN | NaN | 6.489 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| Niue | 5 | NU | New Zealand | NZ | 0.541413 | 0.021414 | POLYGON ((-169.89389 -19.14556, -169.93088 -19... | NaN | NaN | 7.324 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
In [29]:
fig,ax = plt.subplots(1,figsize = (40,24))
sm = plt.cm.ScalarMappable(cmap='viridis',norm = plt.Normalize(vmin = vmin, vmax=vmax))
sm._A = []
cbar = fig.colorbar(sm)
merged.plot(column = 'Happiness Score', cmap = 'viridis', linewidth = 0.3, ax=ax,edgecolor = '0.5')
ax.set_title('World Happiness Score',fontdict = {'fontsize':'40'})
Out[29]:
Text(0.5, 1.0, 'World Happiness Score')
In [30]:
variable='suicide'
fig,ax = plt.subplots(1,figsize = (40,24))
sm = plt.cm.ScalarMappable(cmap='viridis',norm = plt.Normalize(vmin = vmin, vmax=vmax))
sm._A = []
cbar = fig.colorbar(sm)
merged.plot(column = variable, cmap = 'viridis', linewidth = 0.3, ax=ax,edgecolor = '0.8')
ax.set_title('World Suicide rate',fontdict = {'fontsize':'40'})
Out[30]:
Text(0.5, 1.0, 'World Suicide rate')
რელიგიურობა¶
In [31]:
rel_data = pd.read_csv('data/religios_GDP.csv')
rel_data=rel_data.rename(columns={"country":"Country"})
rel_data
Out[31]:
| Unnamed: 0 | Country | religiousity% | US$ | |
|---|---|---|---|---|
| 0 | 0 | Angola | 88 | 4465 |
| 1 | 1 | Brazil | 79 | 9895 |
| 2 | 2 | Bulgaria | 52 | 8077 |
| 3 | 3 | Colombia | 82 | 6379 |
| 4 | 4 | Tanzania | 89 | 1034 |
| ... | ... | ... | ... | ... |
| 143 | 143 | Rwanda | 95 | 772 |
| 144 | 144 | Puerto Rico | 85 | 31581 |
| 145 | 145 | Qatar | 95 | 61024 |
| 146 | 146 | Togo | 80 | 611 |
| 147 | 147 | Taiwan | 45 | 24292 |
148 rows × 4 columns
In [32]:
df_rel = pd.merge(df, rel_data, on=['Country'])
In [33]:
sns.lmplot(y='Happiness Score',x='religiousity%',data=df_rel);
In [34]:
sns.relplot(data=df_rel,x='religiousity%',y='Happiness Score',hue='Region',size="US$",sizes=(5, 400),alpha=0.5);
In [35]:
import plotly.express as px
fig = px.scatter(df_rel, x="religiousity%", y="Happiness Score",color='Region',size='US$',hover_name='Country')
fig.show()
fig.write_html("plotly/file.html")
Covid-ის მონაცემები 1000 000 ადამიანზე¶
In [36]:
covid = pd.read_csv('data/covid.csv')
covid=covid.rename(columns={"Unnamed: 3":"covid"})
covid=covid[['Country','covid']]
covid.dropna(inplace=True)
covid
Out[36]:
| Country | covid | |
|---|---|---|
| 1 | Afghanistan | 1,427.73 |
| 2 | Africa | 2,846.00 |
| 3 | Albania | 34,422.13 |
| 4 | Algeria | 2,548.70 |
| 5 | Andorra | 138,109.11 |
| ... | ... | ... |
| 194 | Vietnam | 24.31 |
| 195 | World | 14,249.67 |
| 196 | Yemen | 72.29 |
| 197 | Zambia | 4,019.37 |
| 198 | Zimbabwe | 2,406.46 |
198 rows × 2 columns
In [37]:
y_2020=y_2020.rename(columns={"Country name":"Country",'Ladder score':'Happiness score'})
In [38]:
cov_data = pd.merge(covid, y_2020, on=['Country'])
In [39]:
cov_data['covid']=cov_data['covid'].str.replace(',', '')
In [66]:
test=cov_data[['covid','Happiness score']];
from sklearn.preprocessing import MinMaxScaler
scaler = MinMaxScaler() ;
scaled_values = scaler.fit_transform(test);
test.loc[:,:] = scaled_values;
C:\Users\zurab\anaconda3\lib\site-packages\pandas\core\indexing.py:1637: SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy C:\Users\zurab\anaconda3\lib\site-packages\pandas\core\indexing.py:692: SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
In [42]:
cov_data['covid']=test['covid']
In [43]:
sns.relplot(data=cov_data,x='Explained by: Log GDP per capita',y='Explained by: Healthy life expectancy',
hue='Regional indicator',size="covid",sizes=(3, 200),alpha=0.5);
წრფივი რეგრესიის მოდელი¶
In [44]:
dum_df = pd.get_dummies(df, columns=["Region"])
In [45]:
X=dum_df.drop(columns=["Country",'Happiness Rank','Trust (Government Corruption)','Standard Error','Happiness Score'])
In [46]:
y=dum_df['Happiness Score']
In [47]:
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=10)
model = LinearRegression()
model.fit(X_train, y_train)
Out[47]:
LinearRegression()
In [48]:
y_pred = model.predict(X_test)
In [49]:
from sklearn import metrics
print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))
print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))
print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))
Mean Absolute Error: 0.09184139883084416 Mean Squared Error: 0.013109356769966943 Root Mean Squared Error: 0.11449609936572924
In [50]:
fig,ax = plt.subplots()
ax.scatter(y_test, y_pred)
ax.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'k--', lw=4)
ax.set_xlabel('Measured')
ax.set_ylabel('Predicted')
fig.show()
<ipython-input-50-73d758206e49>:6: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.
In [51]:
r_sq = model.score(X_train, y_train)
print('coefficient of determination:', r_sq)
coefficient of determination: 0.9940367076549931
In [52]:
suicide_2016=suicide[suicide['Year']==2016]
df_2016 = pd.merge(y_2015, suicide_2015, on=['Country'])
df_2016.drop(columns=['Year'],inplace=True)
dum_df_2016 = pd.get_dummies(df_2016, columns=["Region"])
In [53]:
dum_df.columns
Out[53]:
Index(['Country', 'Happiness Rank', 'Happiness Score', 'Standard Error',
'Economy (GDP per Capita)', 'Family', 'Health (Life Expectancy)',
'Freedom', 'Trust (Government Corruption)', 'Generosity',
'Dystopia Residual', 'suicide', 'Region_Australia and New Zealand',
'Region_Central and Eastern Europe', 'Region_Eastern Asia',
'Region_Latin America and Caribbean',
'Region_Middle East and Northern Africa', 'Region_North America',
'Region_Southeastern Asia', 'Region_Southern Asia',
'Region_Sub-Saharan Africa', 'Region_Western Europe'],
dtype='object')
In [54]:
# X_train=dum_df.drop(columns=["Country",'Happiness Rank','Trust (Government Corruption)','Standard Error','Happiness Score'])
# y_train=dum_df['suicide']
X_test=dum_df_2016.drop(columns=["Country",'Happiness Rank','Trust (Government Corruption)','Standard Error','Happiness Score'])
y_test=dum_df_2016['Happiness Score']
# model_2 = LinearRegression()
# model_2.fit(X_train, y_train)
In [55]:
y_pred = model.predict(X_test)
In [56]:
from sklearn import metrics
print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))
print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))
print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))
Mean Absolute Error: 0.0754315865740917 Mean Squared Error: 0.009057826512209627 Root Mean Squared Error: 0.09517261429744182
In [57]:
fig,ax = plt.subplots()
ax.scatter(y_test, y_pred,alpha=0.3,c='r',s=100)
ax.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'k--', lw=4)
ax.set_xlabel('Measured')
ax.set_ylabel('Predicted')
ax.set_title(' Number of Suicide')
fig.show()
<ipython-input-57-642a7a46ea6b>:7: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.
In [58]:
r_sq = model.score(X_train, y_train)
print('coefficient of determination:', r_sq)
coefficient of determination: 0.9940367076549931
