feat: add both NL and offers

app.py

@@ -11,13 +11,13 @@ load_dotenv(dotenv_path=".env", override=True)
 USER = os.getenv("USERNAME")
 PASS = os.getenv("PASSWORD")
 
-
+# create an interface and limit output's width for the dataframe bu
 list_iface = gr.Interface(fn=compute_offer,
             inputs=[gr.File(label="Upload CSV", type="file"),
                     gr.Slider(1, 365, value=30, step=1, label="Days", info="Number of days to look back"),
                     gr.Slider(5000, 100000, value=15000, step=1, label="Minimum Sent", info="Minimum number of emails sent"),
-                    gr.Dropdown(["Comcast", "Yahoo", "Hotmail", "Aol"], value="Comcast", label="Domain")
-                    ],
+                    gr.Dropdown(["Comcast", "Yahoo", "Hotmail", "Aol"], value="Comcast", label="Domain"),
+                    gr.Radio(["Newsletters", "Offers"], label="Type", value="Newsletters")],
             outputs="dataframe")
 
 # PLOTTING

utils/gradio_utils.py

@@ -1,7 +1,7 @@
 import random
 import ipaddress
 import pandas as pd
-
+import re
 
 ### SWAKS ###
 
@@ -230,36 +230,68 @@ def generate_ips_per_subclass(ip_subclasses: str, num_of_ips: int) -> str:
     return "\n".join(ip_addresses)
 
 
+def _replace_numbers(input_string: str) -> str:
+    # Find the numbers before and inside the parentheses
+    match = re.search(r'(\d+)\s*\((\d+)\)', input_string)
+    if match:
+        # Replace the first set of numbers with the second set
+        replaced_string = input_string.replace(match.group(1), match.group(2), 1)
+        # Remove the parentheses and any surrounding whitespace
+        cleaned_string = re.sub(r'\(\d+\)', '', replaced_string).strip()
+        return cleaned_string
+    else:
+        return input_string
+
+def _limit_chars(input_string: str, limit: int = 35) -> str:
+    return input_string[:limit]
+
 ### GENERATE TOP LISTS ###
-def compute_offer(csv_file, days_lookback, min_sent, domain):
+def compute_offer(csv_file, days_lookback, min_sent, domain, offer_type):
+    pd.set_option('display.max_colwidth', 10)
     # cmp_list = ['MSP', 'HOM', 'NTU', 'HCK', 'DDS', 'MNP', 'PSC', 'DTL', 'GVS', 'ANP', 'WDR', 'BSG']   #1
 
     #raw_df = pd.read_csv('tools/data/30.08.2023.gabriel.sabau.campanii.csv', parse_dates=['Data'])
-    comcast_df = pd.read_csv(csv_file.name, parse_dates=['Data'])
+    df_all = pd.read_csv(csv_file.name, parse_dates=['Data'])
     # raw_df = pd.read_csv(csv_file.name, parse_dates=['Data']) #1
 
     cols = ['Campanie', 'Oferta', 'Nume', 'Server', 'User',
         'Lista Custom', 'Data', 'HClicks', 'Clicks', 'Unscribers', 'Openers',
         'Click Open', 'Leads', 'CLike', 'Complains', 'Traps', 'Send']
-    # comcast_df = raw_df[raw_df['Nume'].str.contains('|'.join(cmp_list))]  #1
-    # comcast_df = comcast_df[comcast_df['Domeniu'] == 'Comcast'] #2
-    exclude_list = comcast_df[(comcast_df['Data'] > (pd.Timestamp('now') - pd.Timedelta(days=days_lookback))) \
-                            & (comcast_df['Domeniu'] == domain)]['Oferta'].unique()
-    comcast_df = comcast_df[~comcast_df['Oferta'].isin(exclude_list)]
+    # df_all = raw_df[raw_df['Nume'].str.contains('|'.join(cmp_list))]  #1
+    # df_all = df_all[df_all['Domeniu'] == 'Comcast'] #2
+    exclude_list = df_all[(df_all['Data'] > (pd.Timestamp('now') - pd.Timedelta(days=days_lookback))) \
+                            & (df_all['Domeniu'] == domain)]['Oferta'].unique()
+    df_all = df_all[~df_all['Oferta'].isin(exclude_list)]
     
-    comcast_df = comcast_df[comcast_df['Send'] > int(min_sent)]
-    comcast_df = comcast_df[cols]
-    comcast_df['Click Open'] = comcast_df['Click Open'].str.replace('%', '').astype(float)
-    
-    # comcast_df = comcast_df[comcast_df['Lista Custom'].str.contains('open')]
-    comcast_df = comcast_df[comcast_df['Nume'].str.startswith("Aeon News") & comcast_df['Nume'].str.contains(r'\b[A-Z]{3}\b.*\b\d{4}\*?\s*(\(\d{4}\))?\b')]
-
-    comcast_df.reset_index(drop=True, inplace=True)
-
-    final_df = comcast_df.groupby(["Oferta", "Nume"])\
+    df_all = df_all[df_all['Send'] > int(min_sent)]
+    df_all = df_all[cols]
+    df_all['Click Open'] = df_all['Click Open'].str.replace('%', '').astype(float)
+    # fixed a blank line in the csv
+    df_all = df_all[df_all["Oferta"] != " "]
+
+    # Limit the characters in the "Nume" column
+    # df_all["Nume"] = df_all["Nume"].apply(_limit_chars)
+
+    # Filter for newsletters or offers
+    if offer_type == "Newsletters":
+        df_all = df_all[df_all['Nume'].str.startswith("Aeon News") & \
+                                (~df_all['Nume'].str.contains(r'\(\d{4}\)')) & \
+                                (df_all['Nume'].str.contains(r' \d{4}$'))]
+    elif offer_type == "Offers":
+        df_all = df_all[~df_all['Nume'].str.startswith("Aeon News")]
+    # Compress the newsletter names
+    # df_all = df_all[df_all['Nume'].str.contains(r'\b[A-Z]{3}\b.*\b\d{4}\*?\s*(\(\d{4}\))?\b')]
+    # df_all['Nume'] = df_all['Nume'].apply(_replace_numbers)
+    # exclude again after the transformation
+    # df_all = df_all[~df_all['Oferta'].isin(exclude_list)]
+
+    df_all.reset_index(drop=True, inplace=True)
+
+    final_df = df_all.groupby(["Oferta", "Nume"])\
         .agg( N=('Oferta', 'count'), send_avg=('Send', 'mean'), CO=('Click Open', 'mean'))\
         .sort_values(['CO', 'N'], ascending=False)
-    final_df['send_avg'] = final_df['send_avg'].round(2).astype(float)
+    final_df['send_avg'] = final_df['send_avg'].astype(int)
+    final_df['CO'] = final_df['CO'].round(2).astype(float)
     final_df.reset_index(inplace=True)
 
     return final_df