[1]. Tool: Google AI Mode
      Prompt: can I call .split() or an equivalent on a series of strings
      Output:
      
      import pandas as pd

      data = {'address': ['123 Main St, Anytown, USA', '456 Oak Ave, Somewhere, CA', '789 Pine Rd, Nowhere, NY']}
      df = pd.DataFrame(data)

      # Split the 'address' column and expand the result into three new columns
      df[['street', 'city', 'state']] = df['address'].str.split(', ', expand=True)

      print(df)

[2]. Tool: Google Gemini
      Prompt: How do I efficiently convert a Series of strings to one of integers based on first letter in each string
      Output:
        import pandas as pd

        s = pd.Series(['Cherry', 'Banana', 'Apple', 'Blueberry', 'Avocado'])

        # Convert the first letters to a category type and get the codes
        codes = s.str[0].astype('category').cat.codes

        print(codes)

[3]. Tool: Google Gemini
      Prompt (In conversation with [2]): I want to specify my own codes
      Output:
        custom_map = {
            'A': 100,
            'B': 250,
            'C': 310
        }
        
        import pandas as pd
        import numpy as np # Used for np.nan

        s = pd.Series(['Apple', 'Banana', 'Cherry', 'Avocado', 'Durian'])

        # Apply the custom map to the first letter of each string
        s_encoded = s.str[0].map(custom_map)

        print(s_encoded)

[4]. Tool: Google Gemini
      Prompt: dtstring = data['DateTime'].str.extract(r'(\d+)')
              data['year'] = dtstring[0].astype(int)
              data['month'] = dtstring[1].astype(int)
              data['date'] = dtstring[2].astype(int)
              data['minute'] = dtstring[3].astype(int) * 60 + dtstring[4].astype(int)

              what's wrong with this?
      
      Output:
        # Regex with multiple capture groups for year, month, day, hour, and minute
        pattern = r'(\d{4}).*?(\d{2}).*?(\d{2}).*?(\d{2}).*?(\d{2})'
        dtstring = data['DateTime'].str.extract(pattern)

        # Now dtstring has 5 columns (0 to 4)
        data['year'] = dtstring[0].astype(int)
        data['month'] = dtstring[1].astype(int)
        data['date'] = dtstring[2].astype(int)
        # Note: The calculation seems to be for the minute of the day
        data['minute_of_day'] = dtstring[3].astype(int) * 60 + dtstring[4].astype(int)

[5]. Tool: Google Gemini
      Prompt: can I just collect all groups of digits?
      Output:
        import pandas as pd

        # Sample data
        data = pd.DataFrame({'DateTime': ['2025-10-06 16:45:30', 'Date:2024/09/01 Time:08-30-00']})

        # 1. Find all groups of digits and get a Series of lists
        digit_lists = data['DateTime'].str.findall(r'(\d+)')
        # This returns:
        # 0    [2025, 10, 06, 16, 45, 30]
        # 1    [2024, 09, 01, 08, 30, 00]
        # Name: DateTime, dtype: object

        # 2. Expand the lists into a DataFrame
        dt_parts = pd.DataFrame(digit_lists.tolist(), index=data.index)

        # 3. Assign columns, converting to integer type
        data['year'] = dt_parts[0].astype(int)
        data['month'] = dt_parts[1].astype(int)
        data['date'] = dt_parts[2].astype(int)
        data['hour'] = dt_parts[3].astype(int)
        data['minute'] = dt_parts[4].astype(int)

[6]. Tool: Google Gemini
      Prompt: x = data[['AgeInDays':'HasName'], ['Type_Cat':'Sex_Unknown']] can I do this?
      Output: 
        import pandas as pd
        import numpy as np

        # Sample DataFrame with non-contiguous column names
        data = pd.DataFrame(np.random.rand(5, 10), columns=[
            'AgeInDays', 'AgeInMonths', 'HasName',
            'Type_Cat', 'Type_Dog', 'Type_Bird',
            'Sex_Male', 'Sex_Female', 'Sex_Unknown',
            'Other_Col'
        ])

        # Create a combined list of columns by extracting columns from the .loc slices
        cols_to_select = data.loc[:, 'AgeInDays':'HasName'].columns.tolist() + \
                        data.loc[:, 'Type_Cat':'Sex_Unknown'].columns.tolist()

        # Select the combined columns using the single list
        x = data[cols_to_select].copy()