[VIZ] tweaking plots

README.md

@@ -11,8 +11,8 @@ Enhanced access to space, climate and natural resource monitoring, space-based a
 This study measures the innovation success of selected companies worldwide based on internal policies, management approaches, and market demand. Time-series metrics, including Return on Equity, Net Profit Margin, Financial Leverage, Asset Turnover, Asset Growth Rate, Equity Growth Rate, and Liabilities Growth Rate, indicate financial trends for businesses across the commercial EO landscape. An Auto-Regressive Distributed Lag model is also implemented to identify the impacts of assets, equity, liabilities, and number of satellites on company revenue and EBITDA. The model is used to forecast revenue and EBITDA into the future for each company.
 
 <p align="center">
-  <img src="images/ebitda_separate.png" width="900" title="EBITDA">
-  <img src="images/revenue_separate.png" width="900" title="Revenue">
+  <img src="images/ebitda_separate.png" width="700" title="EBITDA">
+  <img src="images/revenue_separate.png" width="700" title="Revenue">
 </p>
 
 <p align="center">

src/plot_utils.py

@@ -38,7 +38,7 @@ def plotKPIs(fig, ax, data, key, companyMetadata, company, ylabel):
     return fig
 
 def plotPredictions(fig, ax, company, companyMetadata, time=None, inSamplePrediction=None, outSamplePrediction=None, ground_truth=None, ylabel=None, split=None, plotEverything=False):
-    size = 22
+    size = 32
     if not plotEverything:
         company_color = np.array(companyMetadata[company]["color"]) / 255
         ax.set_title(company, fontsize=16, fontweight='bold')
@@ -56,8 +56,8 @@ def plotPredictions(fig, ax, company, companyMetadata, time=None, inSamplePredic
         ax.set_position([box.x0, box.y0 + box.height * 0.2, box.width, box.height * 0.8])
     elif plotEverything:
         company_color = np.array(companyMetadata[company]["color"]) / 255
-        ax.set_title(company, fontsize=size, fontweight='bold')
-        ax.tick_params(axis='both', which='major', labelsize=18)
+        ax.set_title(company, fontsize=40, fontweight='bold')
+        ax.tick_params(axis='both', which='major', labelsize=size)
         ax.set_xlabel('Year', fontsize=size)
         ax.set_ylabel(ylabel, fontsize=size)
         ax.set_yscale('symlog')
@@ -69,16 +69,16 @@ def plotPredictions(fig, ax, company, companyMetadata, time=None, inSamplePredic
         ax.xaxis.set_major_locator(ticker.MaxNLocator(nbins=5))
         ax.set_ylim([vmin, vmax])
         ax.set_xlim([pd.to_datetime('2018-06-01'), pd.to_datetime('2024-12-31')])
-        ax.plot(time[:-split], inSamplePrediction, color=company_color, linestyle="", marker=".", markersize=18)
-        ax.plot(time[-split:], outSamplePrediction, color=company_color, marker="*", markersize=18, linestyle="")
-        ax.plot(time, ground_truth, color=company_color, linestyle="solid", linewidth=12)
+        ax.plot(time[:-split], inSamplePrediction, color=company_color, alpha=0.7, linestyle="", marker="s", markersize=40)
+        ax.plot(time[-split:], outSamplePrediction, color=company_color, alpha=0.7, marker="*", markersize=40, linestyle="")
+        ax.plot(time, ground_truth, color=company_color, linestyle="solid", linewidth=20)
     if ylabel=="Revenue $ (USD)":
-        line1, = ax.plot([], [], color='k', linestyle="", marker=".", markersize=18, label='In-Sample Prediction')
-        line2, = ax.plot([], [], color='k', marker="*", markersize=18, linestyle="", label='Out-of-Sample Forecast')
-        line3, = ax.plot([], [], color='k', linestyle="solid", linewidth=12, label='Ground Truth Data')
+        line1, = ax.plot([], [], color='k', linestyle="", alpha=0.7, marker="s", markersize=40, label='In-Sample Prediction')
+        line2, = ax.plot([], [], color='k', marker="*", alpha=0.7, markersize=40, linestyle="", label='Out-of-Sample Forecast')
+        line3, = ax.plot([], [], color='k', linestyle="solid", linewidth=20, label='Ground Truth Data')
         # Add the legend with specified labels
         fig.legend([line1, line2, line3], ['In-Sample Prediction', 'Out-of-Sample Forecast', 'Ground Truth Data'],
-            loc='upper center', bbox_to_anchor=(0.5, 0.05), ncol=3, prop={'size': size})
+            loc='upper center', bbox_to_anchor=(0.5, 0.0), ncol=3, prop={'size': size}, fontsize=size)
     return fig
 
 # Fixes setting number of ticks on a symlog scale, copied from:

src/process.py

@@ -162,8 +162,8 @@ def generate_next_quarters(start_date, n):
 fig14, ax14 = plt.subplots(figsize=(10, 7), dpi=300)
 fig15, ax15 = plt.subplots(figsize=(10, 7), dpi=300)
 fig16, ax16 = plt.subplots(figsize=(10, 7), dpi=300)
-fig17, ax17 = plt.subplots(1, 5, figsize=(55, 10), dpi=300)
-fig18, ax18 = plt.subplots(1, 5, figsize=(55, 10), dpi=300)
+fig17, ax17 = plt.subplots(1, 5, figsize=(70, 10), dpi=300)
+fig18, ax18 = plt.subplots(1, 5, figsize=(70, 10), dpi=300)
 
 for i, company in enumerate(companyMetadata):
     print("COMPANY: ", company)
@@ -232,8 +232,6 @@ def generate_next_quarters(start_date, n):
         plot_pacf(transformedData["EBITDA_USD"], lags=2, title=company, ax = ax30[i//3, i%3], color = 'red')
         plotPredictions(fig17, ax17[i], company, companyMetadata, filteredData["Quarter"], inSamplePredictions1, outSamplePredictions1, transformedData["Revenue_USD"], "Revenue $ (USD)", split, plotEverything=True)
         plotPredictions(fig18, ax18[i], company, companyMetadata, filteredData["Quarter"], inSamplePredictions2, outSamplePredictions2, transformedData["EBITDA_USD"], "EBITDA $ (USD)", split, plotEverything=True)
-    else:
-        pass
 
     plot_acf(transformedData["Total_Assets_USD"], lags=3, title=company, ax = ax21[i//3, i%3], color = 'red')
     plot_pacf(transformedData["Total_Assets_USD"], lags=2, title=company, ax = ax22[i//3, i%3], color = 'red')