-

app.py

@@ -124,20 +124,27 @@ def get_most_liked_users():
     return y.iloc[:7].astype(str), fig
 
 
-def get_sdk_proportions():
-    fig = plt.figure()
-    colors = ["#E13F29", "#D69A80", "#D63B59", "#AE5552", "#CB5C3B", "#EB8076", "#96624E"]
-    spaces['sdk'].value_counts().plot.pie(
-        shadow=False,
-        colors=colors,
-        startangle=90,
-        autopct='%1.1f%%')
-
-    plt.axis('equal')
-    plt.tight_layout()
+def pie_plot(data, figsize=(10, 5)):
+    colors = ['#ff9999', '#66b3ff', '#99ff99', '#ffcc99']
+
+    fig = plt.figure(figsize=figsize)
+    for i, (categories, counts) in enumerate(data.items()):
+        plt.subplot(1, len(data), i+1)
+        plt.pie(list(map(int, counts.split(","))), colors=colors, labels=categories.split(","), autopct='%1.1f%%', startangle=90)
+        # draw circle
+        centre_circle = plt.Circle((0, 0), 0.70, fc='white')
+        plt.gcf().gca().add_artist(centre_circle)
+        # Equal aspect ratio ensures that pie is drawn as a circle
+        plt.axis('equal')
+        plt.tight_layout()
     return fig
 
 
+def get_sdk_proportions():
+    df = spaces.groupby('sdk').size().reset_index(name='counts')
+    return pie_plot({",".join(list(df['sdk'])): ",".join(df['counts'].astype(str))})
+
+
 def get_sdk_frequencies():
     fig = plt.figure()
     sns.stripplot(x='sdk', y='likes', data=spaces, jitter=True)
@@ -146,54 +153,43 @@ def get_sdk_frequencies():
 
 
 def get_io_proportions():
-    fig = plt.figure()
-    colors = ["#E13F29", "#D69A80", "#D63B59", "#AE5552", "#CB5C3B", "#EB8076", "#96624E"]
-    a = spaces[spaces['inputs'].notnull()]['inputs'].values
-    a = [y.split(',') for y in a]
-    flat_list = [x for xs in a for x in xs]
-    pd.Series(flat_list).value_counts().plot.pie(
-        shadow=False,
-        colors=colors,
-        # explode=(0.15, 0, 0),
-        startangle=90,
-        autopct='%1.1f%%')
-
-    plt.axis('equal')
-    plt.tight_layout()
-    return fig
+    inputs = [y.split(',') for y in spaces[spaces['inputs'].notnull()]['inputs'].values]
+    inputs = [x for xs in inputs for x in xs]
+    inputs = pd.Series(inputs).value_counts()
+    mask = (inputs/inputs.sum() * 100).lt(2)
+    updated_inputs = inputs[~mask]
+    updated_inputs['Other'] = inputs[mask].sum()
+
+    outputs = [y.split(',') for y in spaces[spaces['outputs'].notnull()]['outputs'].values]
+    outputs = [x for xs in outputs for x in xs]
+    outputs = pd.Series(outputs).value_counts()
+    mask = (outputs/outputs.sum() * 100).lt(2)
+    updated_outputs = outputs[~mask]
+    updated_outputs['Other'] = outputs[mask].sum()
+    
+    return pie_plot({
+        ",".join(list(updated_inputs.index.astype(str))): ",".join(updated_inputs.values.astype(str)),
+        ",".join(list(updated_outputs.index.astype(str))): ",".join(updated_outputs.values.astype(str))
+    })
 
 
 def get_packages_proportions():
-    fig = plt.figure()
-    colors = ["#E13F29", "#D69A80", "#D63B59", "#AE5552", "#CB5C3B", "#EB8076", "#96624E"]
-    a = spaces[spaces['ai_ml_reqs'].notnull()]['ai_ml_reqs'].values
-    a = [y.split(',') for y in a]
-    flat_list = [x for xs in a for x in xs]
-    pd.Series(flat_list).value_counts().plot.pie(
-        shadow=False,
-        colors=colors,
-        # explode=(0.15, 0, 0),
-        startangle=90,
-        autopct='%1.1f%%')
-
-    plt.axis('equal')
-    plt.tight_layout()
-    return fig
+    spaces_ai_reqs = [y.split(',') for y in spaces[spaces['ai_ml_reqs'].notnull()]['ai_ml_reqs'].values]
+    spaces_ai_reqs = [x for xs in spaces_ai_reqs for x in xs]
+    spaces_ai_reqs = pd.Series(spaces_ai_reqs).value_counts()
+    mask = (spaces_ai_reqs/spaces_ai_reqs.sum() * 100).lt(3)
+    updated_spaces_ai_reqs = spaces_ai_reqs[~mask]
+    updated_spaces_ai_reqs['Other'] = spaces_ai_reqs[mask].sum()
+    print(updated_spaces_ai_reqs)
+    return pie_plot({",".join(list(updated_spaces_ai_reqs.index.astype(str))): ",".join(updated_spaces_ai_reqs.values.astype(str))})
 
 
 def get_processable_spaces_proportions():
-    fig = plt.figure()
-    colors = ["#E13F29", "#D69A80", "#D63B59", "#AE5552", "#CB5C3B", "#EB8076", "#96624E"]
-    spaces['is_processed'].value_counts().plot.pie(
-        shadow=False,
-        colors=colors,
-        # explode=(0.15, 0, 0),
-        startangle=90,
-        autopct='%1.1f%%')
-
-    plt.axis('equal')
-    plt.tight_layout()
-    return fig
+    spaces_status = spaces['status'].value_counts()
+    mask = (spaces_status/spaces_status.sum() * 100).lt(5)
+    updated_spaces_status = spaces_status[~mask]
+    updated_spaces_status['Error'] = spaces_status[mask].sum()
+    return pie_plot({",".join(list(updated_spaces_status.index.astype(str))): ",".join(updated_spaces_status.values.astype(str))})
 
 
 def get_tsne():