Forex Forecasting with Random Forests /ALL – Back/forward test and importances

### Back test
# Leverage multiple
leverage = 25
# Balance at start
start_balance = 1000000

# Profit and loss calculation
df_train_result['is_train'] = True
df_test_result['is_train'] = False
df_concat = pd.concat([df_train_result, df_test_result])

df_back_test = trade_simulation(df_concat,leverage,start_balance)
df_back_test.to_csv('/Users/alisa/github/ml/fx/data/output/df_back_test_1.csv', encoding='cp932',index=True)

# Visualization of the balance during backtesting
fig, ax = plt.subplots(figsize=(16, 4))
ax.plot(df_concat[df_back_test['is_train']]['Date'], df_back_test[df_back_test['is_train']]['balance'], label='train_balance')
ax.plot(df_concat[~df_back_test['is_train']]['Date'], df_back_test[~df_back_test['is_train']]['balance'], label='test_balance')
ax.legend(fontsize=12)
ax.set_title('Back test', fontsize=14)
ax.yaxis.get_major_formatter().set_useOffset(False)
ax.yaxis.get_major_formatter().set_scientific(False) 
ax.yaxis.set_major_locator(ticker.MaxNLocator(integer=True))
ax.set_ylim(0, df_back_test['balance'].max() + df_back_test['balance'].max()*0.1)
# ax.yaxis.set_ticks(np.arange(0, df_back_test['balance'].max() + 1000000, 500000))

ax.xaxis.set_major_locator(mdates.MonthLocator(interval=12))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y'))
ax.tick_params(axis='x', labelrotation=45)

print('First day: {}'.format(df_back_test.iloc[0]['Date']))
print('Last day: {}'.format(df_back_test.iloc[-1]['Date']))
print('Balance at last day: {}'.format(df_back_test.iloc[-1]['balance'].astype(int)))
print('profit: {}'.format(df_back_test.iloc[-1]['balance'].astype(int)-start_balance))

## Forward test
# Leverage multiple
leverage = 25
# Balance at start
start_balance = 1000000
# Profit and loss calculation
df_forward_test = trade_simulation(df_test_result,leverage,start_balance)
df_forward_test.to_csv('/Users/alisa/github/ml/fx/data/output/df_forward_test_1.csv', encoding='cp932',index=True)

# Visualization Balance at Forward test
fig, ax = plt.subplots(figsize=(16, 4))
ax.plot(df_forward_test['Date'], df_forward_test['balance'], label='balance', color='orange')
ax.legend(fontsize=12)
ax.set_title('Forward test', fontsize=14)
ax.yaxis.get_major_formatter().set_useOffset(False)
ax.yaxis.get_major_formatter().set_scientific(False) 
ax.yaxis.set_major_locator(ticker.MaxNLocator(integer=True))
ax.set_ylim(0, df_forward_test['balance'].max() + df_forward_test['balance'].max()*0.1)
# ax.yaxis.set_ticks(np.arange(0, 5000001, 500000))

ax.xaxis.set_major_locator(mdates.MonthLocator(interval=2))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%y/%m'))
ax.tick_params(axis='x', labelrotation=45)

print('Start day: {}'.format(df_forward_test.iloc[0]['Date']))
print('Last day: {}'.format(df_forward_test.iloc[-1]['Date']))
print('Balance at last day: {}'.format(df_forward_test.iloc[-1]['balance'].astype(int)))
print('Profit: {}'.format(df_forward_test.iloc[-1]['balance'].astype(int)-start_balance))

### feature_importance
def plot_feature_importance(importance,names):
    
    # Create arrays from feature importance and feature names
    feature_importance = np.array(importance)
    feature_names = np.array(names)

    # Create a DataFrame using a Dictionary
    data={'feature_names':feature_names,'feature_importance':feature_importance}
    fi_df = pd.DataFrame(data)

    # Sort the DataFrame in order decreasing feature importance
    fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)
    return fi_df

# Get the top 20 most important variables
fi_df = plot_feature_importance(model.feature_importances_,X_train.columns)
fi_df = fi_df.head(20)

# Define size of bar plot
plt.figure(figsize=(10,8))
# Plot Searborn bar chart
sns.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])
# Add chart labels
plt.title('RANDOM FOREST' + 'FEATURE IMPORTANCE')
plt.xlabel('FEATURE IMPORTANCE')
plt.ylabel('FEATURE NAMES')

display(fi_df)