import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Set font size
font = {'family' : 'normal',
        'weight' : 'normal',
        'size'   : 16}
plt.rc('font', **font)

# Read data using pandas, as it contains mixed data types (numbers and strings)
data = pd.read_csv('green.csv')

# Create new variable containing only the rows of the database where TEA is the base
tea = data[data['Base']=='TEA']
# Create new variable containing only a list of yields, when TEA was used as the base.
tea_yield = tea['Yield (%)'].values

# As above, but for DIPEA
dipea = data[data['Base']=='DIPEA']
dipea_yield = dipea['Yield (%)'].values

# Calculate the average yield when DIPEA and TEA were used as the base
tea_av = np.average(tea_yield)
dipea_av = np.average(dipea_yield)

# Plot and save the bar chart
plt.bar(['TEA','DIPEA'],[tea_av,dipea_av])
plt.xlabel('Base')
plt.ylabel('Yield (%)')

# Save the figure. bbox_inches sets the borders automatically, and dpi sets the resolution of the image
plt.savefig('bar.png',bbox_inches='tight',dpi=500)