Global Economic Trade Data

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

import seaborn as sns
import plotly.plotly as py

df = pd.read_csv(r"D:\Downloads\commodity_trade_statistics_data.csv\commodity_trade_statistics_data.csv")

df.head()

df.info()

np.unique(df.year)

df.groupby(['year', 'flow'])['trade_usd'].sum().unstack().plot(kind='bar', figsize=(15,10))
plt.ylabel('Trade Value(USD)')
plt.show()

country_flows = df.groupby(['country_or_area', 'flow'])['trade_usd'].sum().unstack()
country_flows.fillna(0, inplace=True)

country_flows['Export'] = country_flows['Export'] + country_flows['Re-Export']
country_flows['Import'] = country_flows['Import'] + country_flows['Re-Import']
country_flows.drop(['Re-Export', 'Re-Import'], axis=1, inplace=True)

country_flows['Delta'] = country_flows['Export'] - country_flows['Import']

df_codes = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/master/2014_world_gdp_with_codes.csv')
df_codes.set_index(['COUNTRY'], inplace=True)

country_flows_merged = country_flows.merge(df_codes, how='left', left_index=True, right_index=True)
country_flows_merged.loc[country_flows_merged.index == 'USA', 'CODE'] = 'USA'
country_flows_merged.loc[country_flows_merged.index == 'China, Hong Kong SAR', 'CODE'] = 'CHN'
country_flows_merged.loc[country_flows_merged.index == 'China, Macao SAR', 'CODE'] = 'CHN'
country_flows_merged.loc[country_flows_merged.index == 'Russian Federation', 'CODE'] = 'RUS'
country_flows_merged.loc[country_flows_merged.index == 'Bahamams', 'CODE'] = 'BHM'
country_flows_merged.loc[country_flows_merged.index == 'Viet Nam', 'CODE'] = 'VNM'
country_flows_merged.loc[country_flows_merged.index == 'Rep. of Korea', 'CODE'] = 'KOR'

country_flows_merged_clean = pd.DataFrame(country_flows_merged.groupby(['CODE'])['Delta'].sum())

data = [ dict(
        type = 'choropleth',
        locations = country_flows_merged_clean.index,
        z = country_flows_merged_clean['Delta'],
        colorscale = 'Viridis',
        autocolorscale = False,
        reversescale = True,
        marker = dict(
            line = dict (
                color = 'rgb(180,180,180)',
                width = 0.5
            ) ),
        colorbar = dict(
            autotick = False,
            tickprefix = '$',
            title = 'US$'),
      ) ]

layout = dict(
    title = 'Total Trade Surplus 1988-2016',
    geo = dict(
        showframe = False,
        showcoastlines = False,
        projection = dict(
            type = 'Mercator'
        )
    )
)

fig = dict( data=data, layout=layout )
py.iplot( fig, validate=False, filename='d3-world-map' )

fig, ax = plt.subplots(1,1)
plt.suptitle('Total Global Trade Surplus 1988-2016', size=20, fontweight='bold')
country_flows['Delta'].sort_values(ascending=False).head(20).plot(kind='barh', figsize=(15,10))
plt.xlabel('Trade Surplus (USD)')
fmt = '${x:,.0f}'
tick = mtick.StrMethodFormatter(fmt)
ax.xaxis.set_major_formatter(tick) 
plt.xticks(rotation=25)
plt.show()

fig, ax = plt.subplots(1,1)

country_flows['Delta'].sort_values(ascending=False).tail(20).plot(kind='barh', figsize=(15,10))
plt.xlabel('Trade Surplus (USD)')
tick = mtick.StrMethodFormatter(fmt)
ax.xaxis.set_major_formatter(tick) 
plt.xticks(rotation=25)
plt.show()

country_ts = df.groupby(['country_or_area', 'year', 'flow'])['trade_usd'].sum().unstack()
country_ts.fillna(0, inplace=True)
country_ts['Export'] = country_ts['Export'] + country_ts['Re-Export']
country_ts['Import'] = country_ts['Import'] + country_ts['Re-Import']
country_ts.drop(['Re-Export', 'Re-Import'], axis=1, inplace=True)
country_ts['Surplus'] = country_ts['Export'] - country_ts['Import']

country_ts = country_ts['Surplus']
country_ts = country_ts.unstack().T
country_ts.fillna(0, inplace=True)

country_ordered = country_ts.mean(axis=0).sort_values().index
top_20 = country_ordered[0:20]
bottom_20 = country_ordered[-20:]

fig, ax = plt.subplots(1,1)
country_ts.loc[:, country_ts.columns.isin(top_20)].plot(figsize=(15,10),ax=ax)
fmt = '${x:,.0f}'
plt.ylabel('Trade Surplus (USD)')
tick = mtick.StrMethodFormatter(fmt)
ax.yaxis.set_major_formatter(tick) 
plt.show()


fig, ax = plt.subplots(1,1)
country_ts.loc[:, country_ts.columns.isin(bottom_20)].plot(figsize=(15,10), ax=ax)
fmt = '${x:,.0f}'
tick = mtick.StrMethodFormatter(fmt)
ax.yaxis.set_major_formatter(tick) 
plt.ylabel('Trade Surplus (USD)')
plt.show()

country_ts

sns.clustermap(country_ts.corr(), figsize=(20,20))
plt.show()

fig, ax = plt.subplots()
category_value = df.loc[df.category != 'all_commodities'].groupby(['category'])['trade_usd'].sum()
category_value.sort_values(ascending=False).plot(kind='barh', figsize=(15,15), ax=ax)
fmt = '${x:,.0f}'
tick = mtick.StrMethodFormatter(fmt)
ax.xaxis.set_major_formatter(tick) 
plt.xticks(rotation=20)
plt.ylabel('Total Trade Value (USD)')
plt.show()

fig, ax = plt.subplots()
category_weight = df.loc[df.category != 'all_commodities'].groupby(['category'])['weight_kg'].sum()
category_weight.sort_values(ascending=False).plot(kind='barh', figsize=(15,15), ax=ax)
fmt = '{x:,.0f}'
tick = mtick.StrMethodFormatter(fmt)
ax.xaxis.set_major_formatter(tick) 
plt.ylabel('Total Trade Weight (kg)')
plt.show()

fig, ax = plt.subplots(1,1)
value_weight = (category_value/category_weight)
value_weight.sort_values(ascending=False).plot(kind='barh', figsize=(15,15))
plt.title('Ratio of Trade Value to Trade Weight', size=20, fontweight='bold')
fmt = '{x:,.0f}'
tick = mtick.StrMethodFormatter(fmt)
ax.xaxis.set_major_formatter(tick) 
plt.xlabel('Trade Value/Trade Weight')
plt.show()

top_valueweight = value_weight.sort_values(ascending=False).index[0:5]

country_category = df.groupby(['country_or_area', 'category'])['trade_usd'].sum().unstack()

country_category = country_category.loc[:, country_category.columns.isin(top_valueweight)]

fig, ax = plt.subplots(3, 2, figsize=(15,10))
for i, j in enumerate(country_category):
    country_category[j].sort_values(ascending=False).head(5).plot(kind='bar', title=j, ax=ax.flat[i])
    ax.flat[i].set_xlabel('')
    
plt.tight_layout()
plt.show()