Upload 4 files (#1)

- Upload 4 files (aa8c619e0bbc53b384a10e20d4edeae9269329d1)

app.py

@@ -0,0 +1,96 @@
+import pandas as pd
+import streamlit as st
+from data_processing import process_dataframe, process_journal, remove_na_accounts
+from utils import get_table_download_link, to_excel
+from io import BytesIO
+from llm_agent import set_openai_key, init_agent, get_agent_response
+
+st.title('Accounting Fast Close')
+
+# Ask user for OpenAI API key
+openai_api_key = st.sidebar.text_input('Enter your OpenAI API Key', type='password')
+if openai_api_key:
+    set_openai_key(openai_api_key)
+
+# Creating a button to toggle between uploaded documents and findings
+view_option = st.selectbox('Choose View', ['Uploaded Documents', 'Findings'])
+
+@st.cache_data
+def load_excel_data(uploaded_file):
+    df = pd.read_excel(uploaded_file)
+    return df
+
+uploaded_file1 = st.sidebar.file_uploader('Upload your trial balance Excel file', type=['xlsx'])
+uploaded_file2 = st.sidebar.file_uploader('Upload your journal entry Excel file', type=['xlsx'])
+
+if uploaded_file1 is not None:
+    df1 = load_excel_data(uploaded_file1)
+    # Process the DataFrame
+    df1 = process_dataframe(df1)
+
+    # AI Agent Section
+    if openai_api_key:
+        agent = init_agent(openai_api_key)
+
+if uploaded_file2 is not None:
+    df2 = load_excel_data(uploaded_file2)
+    df2 = process_journal(df2)
+
+if uploaded_file1 is not None and uploaded_file2 is not None:
+    # Merge df1 (trial balance) with df2 (journal entries) 
+    df1 = pd.merge(df1, df2, on='Account', how='outer')
+
+    # Remove rows with 'Account' as NA
+    df1 = remove_na_accounts(df1)
+
+    # Define the columns we want to fill NaN values with 0
+    fillna_columns = ['Opening Balance Debit', 'Opening Balance Credit', 
+                      'Current Transactions Debit', 'Current Transactions Credit', 
+                      'Closing Balance Debit', 'Closing Balance Credit', 
+                      'Debit Amount', 'Credit Amount' ]
+
+    # Replace NaN values with 0 in the defined columns
+    df1[fillna_columns] = df1[fillna_columns].fillna(0)
+
+    # Compute the differences
+    df1['Diff Dr.'] = df1['Current Transactions Debit'] - df1['Debit Amount']
+    df1['Diff Cr.'] = df1['Current Transactions Credit'] - df1['Credit Amount']
+
+    excel_data = to_excel(df1)  # Move this line to here
+
+# Uploaded Documents Section
+if view_option == 'Uploaded Documents':
+    if uploaded_file1 is not None:
+        st.write(df1)
+#       st.markdown(get_table_download_link(excel_data, 'processed_data.xlsx'), unsafe_allow_html=True)
+
+    if uploaded_file2 is not None:
+        # Save the dataframes to an Excel file
+        excel_data_combined = BytesIO()
+        with pd.ExcelWriter(excel_data_combined, engine='xlsxwriter') as writer:
+            df1.to_excel(writer, sheet_name='Trial Balance', index=False)
+            df2.to_excel(writer, sheet_name='Journal Entry', index=False)
+
+        st.markdown(get_table_download_link(excel_data_combined.getvalue(), filename='combined.xlsx'), unsafe_allow_html=True)
+
+elif view_option == 'Findings':
+    # Logic for findings should be implemented here
+    st.write(df1)
+
+    # AI Agent Section
+    if openai_api_key and uploaded_file1 is not None:
+        # Create a chat box for user questions
+        user_input = st.text_input('Ask a question:')
+        if user_input:
+            response = get_agent_response(agent, df1, user_input)  # pass df1 as an argument
+            st.write(response)
+
+    if uploaded_file2 is not None:
+        # Save the dataframes to an Excel file
+        excel_data_combined = BytesIO()
+        with pd.ExcelWriter(excel_data_combined, engine='xlsxwriter') as writer:
+            df1.to_excel(writer, sheet_name='Trial Balance', index=False)
+            df2.to_excel(writer, sheet_name='Journal Entry', index=False)
+
+    st.markdown(get_table_download_link(excel_data_combined.getvalue(), filename='combined.xlsx'), unsafe_allow_html=True)
+

data_processing.py

@@ -0,0 +1,67 @@
+import pandas as pd
+from typing import List
+
+def remove_nulls(df: pd.DataFrame, columns: List[str]) -> pd.DataFrame:
+    for column in columns:
+        df = df[df[column].notnull() & df[column].astype(str).str[0].str.isdigit()]
+    return df
+
+def remove_na_accounts(df: pd.DataFrame) -> pd.DataFrame:
+    df = df.dropna(subset=['Account'])
+    return df
+
+def remove_empty_columns(df: pd.DataFrame) -> pd.DataFrame:
+    df = df.dropna(how='all', axis=1)
+    return df
+
+def handle_unknown_columns(df: pd.DataFrame) -> pd.DataFrame:
+    df = df.apply(lambda x: x[x.astype(str).str[0].str.isdigit()] if x.dtype in ['object', 'int64'] else x)
+    return df
+
+def rename_columns(df: pd.DataFrame) -> pd.DataFrame:
+    if len(df.columns) == 10:
+        df.columns = ['Account', 'Description', 'Opening Balance Debit', 'Opening Balance Credit',  
+                      'Current Transactions Debit', 'Current Transactions Credit',
+                      'Total Transactions Debit', 'Total Transactions Credit', 
+                      'Closing Balance Debit', 'Closing Balance Credit']
+    elif len(df.columns) == 8:
+        df.columns = ['Account', 'Description', 'Opening Balance Debit', 'Opening Balance Credit', 
+                      'Current Transactions Debit', 'Current Transactions Credit', 
+                      'Closing Balance Debit', 'Closing Balance Credit']
+    return df
+
+def convert_to_float(df: pd.DataFrame, skip_columns: List[str]) -> pd.DataFrame:
+    df = df.apply(lambda x: x.astype(str).str.replace(',', '').astype(float) if x.name not in skip_columns else x)
+    return df
+
+def process_dataframe(df: pd.DataFrame, *args) -> pd.DataFrame:
+    df = remove_nulls(df, args)
+    df = remove_empty_columns(df)
+    df = handle_unknown_columns(df)
+    df = rename_columns(df)
+    df = convert_to_float(df, ['Account', 'Description'])
+    return df
+
+
+def rename_columns_je(df):
+    column_mapping = {
+        'Cont debitor': 'Account Debit',
+        'Cont creditor': 'Account Credit',
+        'Suma': 'Amount' 
+    }
+    
+    df.rename(columns=column_mapping, inplace=True)
+    return df
+
+def process_journal(df: pd.DataFrame) -> pd.DataFrame:
+    df = rename_columns_je(df)
+
+    transactions_dr = df.groupby('Account Debit').agg({'Amount': 'sum'}).reset_index().rename(columns={'Amount': 'Debit Amount', 'Account Debit': 'Account'})
+    transactions_cr = df.groupby('Account Credit').agg({'Amount': 'sum'}).reset_index().rename(columns={'Amount': 'Credit Amount', 'Account Credit': 'Account'})
+
+    df_out = pd.merge(transactions_dr, transactions_cr, on='Account', how='outer')
+
+    df_out.fillna(0, inplace=True)
+    
+    return df_out
+

llm_agent.py

@@ -0,0 +1,18 @@
+import os
+import pandas as pd
+from pandasai import PandasAI
+from pandasai.llm.openai import OpenAI
+
+def set_openai_key(api_key):
+    os.environ["OPENAI_API_KEY"] = api_key
+    return api_key
+
+def init_agent(api_key):
+    api_key = set_openai_key(api_key)
+    llm = OpenAI(api_token=api_key)
+    pandas_ai = PandasAI(llm, conversational=False)
+    return pandas_ai
+
+def get_agent_response(agent, df, user_input):
+    response = agent.run(df, user_input)
+    return response

utils.py

@@ -0,0 +1,44 @@
+import base64
+import pandas as pd
+from io import BytesIO
+import os
+
+def get_table_download_link(excel_data: bytes, filename: str = 'data.xlsx') -> str:
+    """Generates a link allowing the data in a given pandas dataframe to be downloaded
+    in:  dataframe
+    out: href string
+    """
+    b64 = base64.b64encode(excel_data).decode()  # some strings <-> bytes conversions necessary here
+    href = f'<a href="data:application/octet-stream;base64,{b64}" download={filename}>Download Excel File</a>'
+    return href
+
+def convert_and_save_as_csv(uploaded_file) -> str:
+    # Check if the uploaded file is not None
+    if uploaded_file is not None:
+        try:
+            # Read the file with pandas
+            df = pd.read_excel(uploaded_file)
+            # Save the file as a CSV file
+            csv_file_path = os.path.splitext(uploaded_file.name)[0] + '.csv'
+            df.to_csv(csv_file_path, index=False)
+            return csv_file_path
+        except Exception as e:
+            print("Error: ", e)
+            return None
+    else:
+        print("No file uploaded.")
+        return None
+
+def to_excel(df):
+    output = BytesIO()
+    with pd.ExcelWriter(output, engine='openpyxl') as writer:
+        df.to_excel(writer, sheet_name='Sheet1', index=False)
+    return output.getvalue()
+
+def load_data(file_path: str) -> pd.DataFrame:
+    try:
+        df = pd.read_csv(file_path)
+        return df
+    except FileNotFoundError as e:
+        print(f"File not found: {e}")
+        return None