data-cleaning.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":7901372,"sourceType":"datasetVersion","datasetId":4619135}],"dockerImageVersionId":30664,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"!pip install openai==1.14.2","metadata":{"_kg_hide-input":true,"_kg_hide-output":true,"execution":{"iopub.status.busy":"2024-03-22T11:24:39.256400Z","iopub.execute_input":"2024-03-22T11:24:39.256942Z","iopub.status.idle":"2024-03-22T11:24:55.142852Z","shell.execute_reply.started":"2024-03-22T11:24:39.256897Z","shell.execute_reply":"2024-03-22T11:24:55.141802Z"},"trusted":true},"execution_count":5,"outputs":[{"name":"stdout","text":"Requirement already satisfied: openai==1.14.2 in /opt/conda/lib/python3.10/site-packages (1.14.2)\nRequirement already satisfied: anyio<5,>=3.5.0 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (4.2.0)\nRequirement already satisfied: distro<2,>=1.7.0 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (1.9.0)\nRequirement already satisfied: httpx<1,>=0.23.0 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (0.27.0)\nRequirement already satisfied: pydantic<3,>=1.9.0 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (2.5.3)\nRequirement already satisfied: sniffio in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (1.3.0)\nRequirement already satisfied: tqdm>4 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (4.66.1)\nRequirement already satisfied: typing-extensions<5,>=4.7 in /opt/conda/lib/python3.10/site-packages (from openai==1.14.2) (4.9.0)\nRequirement already satisfied: idna>=2.8 in /opt/conda/lib/python3.10/site-packages (from anyio<5,>=3.5.0->openai==1.14.2) (3.6)\nRequirement already satisfied: exceptiongroup>=1.0.2 in /opt/conda/lib/python3.10/site-packages (from anyio<5,>=3.5.0->openai==1.14.2) (1.2.0)\nRequirement already satisfied: certifi in /opt/conda/lib/python3.10/site-packages (from httpx<1,>=0.23.0->openai==1.14.2) (2024.2.2)\nRequirement already satisfied: httpcore==1.* in /opt/conda/lib/python3.10/site-packages (from httpx<1,>=0.23.0->openai==1.14.2) (1.0.4)\nRequirement already satisfied: h11<0.15,>=0.13 in /opt/conda/lib/python3.10/site-packages (from httpcore==1.*->httpx<1,>=0.23.0->openai==1.14.2) (0.14.0)\nRequirement already satisfied: annotated-types>=0.4.0 in /opt/conda/lib/python3.10/site-packages (from pydantic<3,>=1.9.0->openai==1.14.2) (0.6.0)\nRequirement already satisfied: pydantic-core==2.14.6 in /opt/conda/lib/python3.10/site-packages (from pydantic<3,>=1.9.0->openai==1.14.2) (2.14.6)\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# Import","metadata":{}},{"cell_type":"code","source":"import io\nimport re\nimport os\nimport json\nimport base64\nimport warnings\n\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\n\nfrom openai import OpenAI\n\nwarnings.simplefilter('ignore')","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:06.077462Z","iopub.execute_input":"2024-03-22T11:25:06.077932Z","iopub.status.idle":"2024-03-22T11:25:06.085405Z","shell.execute_reply.started":"2024-03-22T11:25:06.077896Z","shell.execute_reply":"2024-03-22T11:25:06.084257Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"markdown","source":"# Constant","metadata":{}},{"cell_type":"code","source":"# Env variables\nos.environ[\"OPEN_API_KEY\"] = \"<OPEN-API-KEY>\"\n\n# File Paths\nCLEANED_DATA_PATH = '/kaggle/input/rta-dubai/sample_clean_application_data_2.xlsx'\nUNCLEANED_DATA_PATH = '/kaggle/input/rta-dubai/sample_unclean_application_data_v2.xlsx'\nMETADATA_PATH = '/kaggle/input/rta-dubai/Application dataset.json'\n\n# Chat-GPT Constants\nTEMP = 1.0\nTOP_P = 1.0\nMAX_TOKENS = 500\nFREQ_PENALTY = 0.5\nPRES_PENALTY = 0.5\nOPEN_AI_CLIENT = OpenAI(api_key=os.getenv(\"OPEN_API_KEY\") ,organization='org-2EugvoZZidKLd5DkFH3dGTIA')","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:06.235333Z","iopub.execute_input":"2024-03-22T11:25:06.235803Z","iopub.status.idle":"2024-03-22T11:25:06.259484Z","shell.execute_reply.started":"2024-03-22T11:25:06.235768Z","shell.execute_reply":"2024-03-22T11:25:06.258271Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"markdown","source":"# Models","metadata":{}},{"cell_type":"code","source":"class MetaData:\n    def __init__(self, data):\n        self.__dict__ = data","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:06.400484Z","iopub.execute_input":"2024-03-22T11:25:06.400926Z","iopub.status.idle":"2024-03-22T11:25:06.406408Z","shell.execute_reply.started":"2024-03-22T11:25:06.400893Z","shell.execute_reply":"2024-03-22T11:25:06.405149Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"markdown","source":"# Helper Function","metadata":{}},{"cell_type":"code","source":"def load_json_and_map_to_class(json_path):\n    try:\n        with open(json_path, 'r') as json_file:\n            data = json.load(json_file)\n            return MetaData(data)\n    except json.JSONDecodeError:\n        print(f\"Error: Unable to decode JSON from file '{file_path}'. File may be empty or not in valid JSON format.\")\n        return None\n\n\ndef extract_column_attributes_for_column_name_mapping(metadata_json):\n    columns_meta_data = []\n    for column in metadata_json.attributes:\n        columns_meta_data.append({\n            \"name\": column['attribute_name'],\n            \"description\": column['description'], \n            \"synonyms\": column['synonyms']\n        })\n    return json.dumps(columns_meta_data)\n\n\ndef convert_dataframe_to_html(dataframe):\n    html_string = dataframe.to_html()\n    return html_string.replace('\\n', '')\n\n\ndef extract_python_code(sample_string):\n    # Define a regular expression pattern to match Python code blocks\n    pattern = r'```python\\n(.*?)```'\n\n    # Find all matches of the pattern in the sample string\n    matches = re.findall(pattern, sample_string, re.DOTALL)\n\n    # Return the Python code blocks found\n    return matches\n\n\ndef convert_column_types_and_map_values_wrt_metadata(metadata_json, dataframe):\n    for column in metadata_json.attributes:\n        if column['attribute_name'] in dataframe.columns:\n            if column['data_type'].lower() == \"string\":\n                dataframe[column[\"attribute_name\"]] = dataframe[column[\"attribute_name\"]].astype(str)\n            elif column['data_type'].lower() == \"enum\":\n                # Check if enum value transformation required\n                if column['is_transformation_required']:\n                    dataframe[column[\"attribute_name\"]] = dataframe[column[\"attribute_name\"]].astype(str).str.lower().map(column['values'])\n                dataframe[column[\"attribute_name\"]] = dataframe[column[\"attribute_name\"]].astype('category')\n            elif column['data_type'].lower() == \"date\":\n                dataframe[column[\"attribute_name\"]] = dataframe[column[\"attribute_name\"]].astype('datetime64[ns]')\n                \n    return dataframe\n\n\ndef check_if_any_new_column_formation_is_required(dataframe, metadata_json):\n    new_columns_data = []\n    for column in metadata_json.attributes:\n        if (column['attribute_name'] not in df_uncleaned) and column['constraints']:\n            column_name = column['attribute_name']\n            column_type = column['data_type']\n            column_desc = column['constraints'][0]\n            new_columns_data.append((column_name, column_type, column_desc))\n    return new_columns_data\n\n\ndef extract_df_column_names(dataframe):\n    return \", \".join(dataframe.columns.tolist())\n\n\ndef extract_df_column_names_with_datatypes(dataframe):\n    columns_data_types = {col: str(dtype) for col, dtype in dataframe.dtypes.items()}\n    json_data = json.dumps(columns_data_types)\n    return json_data\n\n\ndef execute_python_script(python_script):\n    slice_script = python_script[9:-3]\n    exec(slice_script)\n    \n\ndef read_saved_dataframe_output():\n    with open('output.txt', 'r') as file:\n        html_content = file.read()\n\n    # Parse HTML content and convert to DataFrame\n    # [TODO] Later remove this, as only for notebook perspective\n    return pd.read_html(html_content)[0]\n\n\ndef read_saved_plot_output():\n    with open('output.png', 'rb') as img_file:\n        img_base64 = base64.b64encode(img_file.read()).decode('utf-8')\n    return img_base64\n\n\ndef read_saved_dataframe_new_col_output():\n    with open('dataframe.txt', 'r') as file:\n        html_content = file.read()\n\n    # Parse HTML content and convert to DataFrame\n    # [TODO] Later remove this, as only for notebook perspective\n    return pd.read_html(html_content)[0]\n\n\ndef contains_visual_keywords(text):\n    visual_keywords = ['plot', 'graph', 'chart', 'visualization']\n    pattern = re.compile(r'\\b(?:' + '|'.join(visual_keywords) + r')\\b', flags=re.IGNORECASE)\n    return True if re.search(pattern, text) else False\n\n\ndef get_html_tag_for_base64_image(img_base64):\n    return f'<img src=\"data:image/png;base64,{img_base64}\" style=\"max-width: 100%; max-height: 100%;\" alt=\"output\">'\n\n    \ndef show_saved_output_img():\n    img = mpimg.imread('output.png')\n    img_height, img_width, _ = img.shape\n    aspect_ratio = img_width / img_height\n    fig_width = 10  # Set a default width\n    fig_height = fig_width / aspect_ratio\n    fig = plt.figure(figsize=(fig_width, fig_height))\n    plt.imshow(img, interpolation='nearest')\n    plt.axis('off')\n    plt.tight_layout()\n    # return fig  # Return the figure object","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:06.909948Z","iopub.execute_input":"2024-03-22T11:25:06.910405Z","iopub.status.idle":"2024-03-22T11:25:06.943276Z","shell.execute_reply.started":"2024-03-22T11:25:06.910367Z","shell.execute_reply":"2024-03-22T11:25:06.941561Z"},"trusted":true},"execution_count":17,"outputs":[]},{"cell_type":"markdown","source":"# Open AI Utils","metadata":{}},{"cell_type":"code","source":"class RTA_Bot:\n    _instance = None\n\n    def __new__(cls, *args, **kwargs):\n        if not cls._instance:\n            cls._instance = super().__new__(cls, *args, **kwargs)\n        return cls._instance\n\n    @staticmethod\n    def ask(prompt):\n        return OPEN_AI_CLIENT.chat.completions.create(\n            model=\"gpt-3.5-turbo\",\n            messages=[\n            {\n                \"role\": \"user\",\n                \"content\": prompt,\n            }],\n        ).choices[0].message.content","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:07.331383Z","iopub.execute_input":"2024-03-22T11:25:07.331840Z","iopub.status.idle":"2024-03-22T11:25:07.340698Z","shell.execute_reply.started":"2024-03-22T11:25:07.331808Z","shell.execute_reply":"2024-03-22T11:25:07.339131Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"def map_column_names_wrt_meta_data(dataset_column_names, metadata_json):\n    prompt = f\"\"\"\nRole: Act as a data analyst whose job is to help in transforming the dataset.\nTask: Consider below defined \"dataset_column_names\". Look up the \"columns_meta_data_json\" and pick relevant column attribute_name w.r.t to the attribute description and synonyms context. Output should be a json with key as 'dataset_column_names' and value as attribute_name picked from columns_meta_data_json\n\ndataset_column_names: {dataset_column_names}\n\ncolumns_meta_data_json: \n{extract_column_attributes_for_column_name_mapping(metadata_json)}\n\nOutput: Return output in JSON Format defined under sample output.\n\nSample Output: {{\"Column1 Old Name\": \"Column1 New Name\", \"Column2 Old Name\": \"Column2 New Name\", ...}}\n\"\"\"\n    response = RTA_Bot.ask(prompt)\n    return json.loads(response)\n\n\ndef ask_rta_bot(question,dataframe):\n    column_names = extract_df_column_names_with_datatypes(dataframe)\n    \n    if contains_visual_keywords(question):\n        execute_visual_prompt(question,column_names)\n    else:\n        output_dataframe = execute_dataframe_prompt(question,column_names)\n        return output_dataframe\n\n\ndef execute_dataframe_prompt(question, column_names):\n    dataframe_name = 'df_uncleaned'\n    prompt = f\"\"\" \nRole: Act as a data analyst whose job is to extract insights from the dataset based on the users asked questions.\n\nTask: Write a valid Python script for the below-mentioned \"user_question\". Use pre-defined / pre-loaded  \"{dataframe_name}\" only for querying the data.\n\nuser_question: {question}\n\nexisting pandas columns names and their types: {column_names}\n\noutput: The output should only be a valid python script\n\nScript steps:\n1. Code for the task\n2. Form a pandas dataframe of the final output\n3. Then convert output dataframe into html string\n4. Save the output into a file named 'output.txt'\n\"\"\"\n    try:\n        # Generate response\n        python_script = RTA_Bot.ask(prompt)\n\n        # Execute script\n        execute_python_script(python_script)\n\n        # Read output dataframe as html table tag\n        output = read_saved_dataframe_output()\n\n        # Remove file\n        # os.remove('output.txt')\n\n        # Return output\n        return output\n    except Exception as e:\n        print(\"I apologize, but I'm currently unable to fulfill your request. Could you please try again or provide me more details to assist me in better understanding your request?\")\n        return None\n    \n    \ndef execute_visual_prompt(question, column_names):\n    dataframe_name = 'df_uncleaned'\n    prompt = f\"\"\" \nRole: Act as a data analyst whose job is extract insights from the dataset based on the users asked question.\n\nTask: Write a valid python script for the below mentioned \"user_question\". Use pre-defined / pre-loaded {dataframe_name} only for querying the data.\n\nuser_question: {question}\n\nexisting pandas columns names and their types: {column_names}\n\noutput: The output should only be a valid python script\n\nScript steps:\n1. Code for the user's asked question.\n2. Form a plot as the final output but don't write graph plotting code.\n3. Save the output plot as a .png file named 'output.png'\n\"\"\"\n    try:\n        # Generate response\n        python_script = RTA_Bot.ask(prompt)\n\n        # Execute script\n        execute_python_script(python_script)\n\n        # Read output image as html image tag\n        # output = get_html_tag_for_base64_image(read_saved_plot_output())\n        \n        # [REMOVE]\n        # show_saved_output_img()\n        \n        # Remove file\n        # os.remove('output.png')\n\n        # Return output\n        return None # return output\n        \n    except Exception as e:\n        print(\"I apologize, but I'm currently unable to fulfill your request. Could you please try again or provide me more details to assist me in better understanding your request?\")\n        return None\n\ndef execute_create_new_column_prompt(new_column, existing_df_columns_metadata):\n    dataframe_name = 'df_uncleaned'\n    new_column_name = new_column[0]\n    new_column_type = new_column[1]\n    new_column_desc = new_column[2]\n    \n    prompt = f\"\"\"\nRole: Act as a data analyst whose job is to create new columns from the below defined data columns for the existing dataset.\nTask: Write a valid python script for the below mentioned \"user_question\". Use existing pandas \"{dataframe_name}\" only for querying the data.\n\nuser_question: Create a new column named \"{new_column_name}\" with type {new_column_type} and constraints as {new_column_desc}. The new column should directly not merge in \"{dataframe_name}\" but it should be stored in a separate dataframe.\n\nexisting pandas columns names and their types: {existing_df_columns_metadata}\ndataframe_name: {dataframe_name}\n\nScript steps:\n1. Code for the task.\n2. Form a pandas dataframe of the final output with newly added column and its supporting column.\n3. Then convert output dataframe into html string\n4. Save the output into a file named 'dataframe.txt'\n\n    \"\"\"\n    try:\n        # Generate response\n        python_script = RTA_Bot.ask(prompt)\n\n        # Execute script\n        execute_python_script(python_script)\n\n        # Read output dataframe as html table tag\n        output = read_saved_dataframe_new_col_output()\n\n        # Remove file\n        # os.remove('dataframe.txt')\n\n        # Return output\n        return output\n    except Exception as e:\n        print(\"I apologize, but I'm currently unable to fulfill your request. Could you please try again or provide me more details to assist me in better understanding your request?\")\n        return None\n","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:07.534774Z","iopub.execute_input":"2024-03-22T11:25:07.535467Z","iopub.status.idle":"2024-03-22T11:25:07.554138Z","shell.execute_reply.started":"2024-03-22T11:25:07.535428Z","shell.execute_reply":"2024-03-22T11:25:07.552514Z"},"trusted":true},"execution_count":19,"outputs":[]},{"cell_type":"markdown","source":"# Pipeline\n\n1. Extract **unclean dataset column names** and **attributes objects** from the meta-data.\n2. Use Open-AI to map **column new names** w.r.t to the provided meta-data.\n3. Change columns names.\n4. Transform columns types and their values w.r.t to the **metadata**\n5. Check if new columns formation is left w.r.t to the **metadata**\n6. Talk with the transformed dataset","metadata":{}},{"cell_type":"markdown","source":"# Code","metadata":{}},{"cell_type":"code","source":"df_uncleaned = pd.read_excel(UNCLEANED_DATA_PATH)\ndf_uncleaned.info()","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:08.411091Z","iopub.execute_input":"2024-03-22T11:25:08.411632Z","iopub.status.idle":"2024-03-22T11:25:08.584443Z","shell.execute_reply.started":"2024-03-22T11:25:08.411589Z","shell.execute_reply":"2024-03-22T11:25:08.583480Z"},"trusted":true},"execution_count":20,"outputs":[{"name":"stdout","text":"<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 100 entries, 0 to 99\nData columns (total 26 columns):\n #   Column                    Non-Null Count  Dtype         \n---  ------                    --------------  -----         \n 0   Name                      100 non-null    object        \n 1   Short Name                100 non-null    object        \n 2   Description               100 non-null    object        \n 3   Business Owner            100 non-null    object        \n 4   Technical Owner           100 non-null    object        \n 5   # users                   56 non-null     object        \n 6   RPO                       26 non-null     object        \n 7   RTO                       27 non-null     object        \n 8   start date                57 non-null     datetime64[ns]\n 9   end date                  58 non-null     datetime64[ns]\n 10  SLA                       33 non-null     object        \n 11  Vendor                    100 non-null    object        \n 12  Status                    88 non-null     object        \n 13  AGF score                 93 non-null     object        \n 14  DR TYpe                   28 non-null     object        \n 15  Required Availability     45 non-null     object        \n 16  location                  62 non-null     object        \n 17  I/X                       84 non-null     object        \n 18  Type of System (Gartner)  68 non-null     object        \n 19  Across RTA?               84 non-null     object        \n 20  commercial or customized  100 non-null    object        \n 21  Recommendation            100 non-null    object        \n 22  Cloud Migration plan      100 non-null    object        \n 23  Type                      100 non-null    object        \n 24  Mobile compatible         78 non-null     object        \n 25  support many langs        100 non-null    object        \ndtypes: datetime64[ns](2), object(24)\nmemory usage: 20.4+ KB\n","output_type":"stream"}]},{"cell_type":"code","source":"metadata_json = load_json_and_map_to_class(METADATA_PATH)\nmetadata_json","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:08.649545Z","iopub.execute_input":"2024-03-22T11:25:08.650660Z","iopub.status.idle":"2024-03-22T11:25:08.657681Z","shell.execute_reply.started":"2024-03-22T11:25:08.650615Z","shell.execute_reply":"2024-03-22T11:25:08.656531Z"},"trusted":true},"execution_count":21,"outputs":[{"execution_count":21,"output_type":"execute_result","data":{"text/plain":"<__main__.MetaData at 0x790930612b30>"},"metadata":{}}]},{"cell_type":"markdown","source":"### 1. Conversion of un-cleaned data column names w.r.t to the metadata context\n1. Extract **unclean dataset column names** and **attributes objects** from the meta-data.\n2. Use Open-AI to map **column new names** w.r.t to the provided meta-data.\n3. Change columns names.","metadata":{}},{"cell_type":"code","source":"# Extract old column names\nuncleaned_column_old_names = ','.join(df_uncleaned.columns)\n\n# Use Open-Ai to map new column names w.r.t to context\nuncleaned_column_new_names = map_column_names_wrt_meta_data(uncleaned_column_old_names, metadata_json)\n\n# Rename columns\ndf_uncleaned.rename(columns=uncleaned_column_new_names, inplace=True)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:20.288341Z","iopub.execute_input":"2024-03-22T11:25:20.288783Z","iopub.status.idle":"2024-03-22T11:25:30.621142Z","shell.execute_reply.started":"2024-03-22T11:25:20.288749Z","shell.execute_reply":"2024-03-22T11:25:30.619574Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"code","source":"uncleaned_column_new_names","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:30.623237Z","iopub.execute_input":"2024-03-22T11:25:30.623698Z","iopub.status.idle":"2024-03-22T11:25:30.632974Z","shell.execute_reply.started":"2024-03-22T11:25:30.623660Z","shell.execute_reply":"2024-03-22T11:25:30.631323Z"},"trusted":true},"execution_count":25,"outputs":[{"execution_count":25,"output_type":"execute_result","data":{"text/plain":"{'Name': 'Name',\n 'Short Name': 'Short Name',\n 'Description': 'Description',\n 'Business Owner': 'Business Owner',\n 'Technical Owner': 'Technical Owner',\n 'Number of Users': 'Number of Users',\n 'Recovery Point Objective': 'Recovery Point Objective',\n 'Recovery Time Objective': 'Recovery Time Objective',\n 'Go Live Date': 'Go Live Date',\n 'End of Support Date': 'End of Support Date',\n 'SLA Type': 'SLA Type',\n 'Vendor': 'Vendor',\n 'Status': 'Status',\n 'AGF Classification': 'AGF Classification',\n 'DR Type': 'DR Type',\n 'Required Availability': 'Required Availability',\n 'System Hosting Place': 'System Hosting Place',\n 'User Community': 'User Community',\n 'Type of System': 'Type of System',\n 'Enterprise Level': 'Enterprise Level',\n 'Development Type': 'Development Type',\n 'Recommendation': 'Recommendation',\n 'Cloud Migration Strategy': 'Cloud Migration Strategy',\n 'Architecture Type': 'Architecture Type',\n 'Mobile Compliance': 'Mobile Compliance',\n 'Multi Language Support': 'Multi Language Support'}"},"metadata":{}}]},{"cell_type":"markdown","source":"### 2. Convert column types w.r.t to the metadata defined attribute type\nTransform columns types and their values w.r.t to the **metadata**","metadata":{}},{"cell_type":"code","source":"df_uncleaned = convert_column_types_and_map_values_wrt_metadata(metadata_json, df_uncleaned)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:34.459145Z","iopub.execute_input":"2024-03-22T11:25:34.459563Z","iopub.status.idle":"2024-03-22T11:25:34.487015Z","shell.execute_reply.started":"2024-03-22T11:25:34.459530Z","shell.execute_reply":"2024-03-22T11:25:34.486022Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"code","source":"df_uncleaned.dtypes","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:34.680145Z","iopub.execute_input":"2024-03-22T11:25:34.680607Z","iopub.status.idle":"2024-03-22T11:25:34.691786Z","shell.execute_reply.started":"2024-03-22T11:25:34.680568Z","shell.execute_reply":"2024-03-22T11:25:34.690135Z"},"trusted":true},"execution_count":27,"outputs":[{"execution_count":27,"output_type":"execute_result","data":{"text/plain":"Name                                object\nShort Name                          object\nDescription                         object\nBusiness Owner                    category\nTechnical Owner                   category\nNumber of Users                     object\nRecovery Point Objective            object\nRecovery Time Objective             object\nGo Live Date                datetime64[ns]\nEnd of Support Date         datetime64[ns]\nSLA Type                          category\nVendor                            category\nStatus                            category\nAGF Classification                category\nDR Type                           category\nRequired Availability             category\nSystem Hosting Place              category\nUser Community                    category\nType of System                    category\nEnterprise Level                  category\nDevelopment Type                  category\nRecommendation                    category\nCloud Migration Strategy          category\nArchitecture Type                 category\nMobile Compliance                 category\nMulti Language Support            category\ndtype: object"},"metadata":{}}]},{"cell_type":"code","source":"df_uncleaned.head(3)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:36.730760Z","iopub.execute_input":"2024-03-22T11:25:36.731197Z","iopub.status.idle":"2024-03-22T11:25:36.772482Z","shell.execute_reply.started":"2024-03-22T11:25:36.731166Z","shell.execute_reply":"2024-03-22T11:25:36.771551Z"},"trusted":true},"execution_count":28,"outputs":[{"execution_count":28,"output_type":"execute_result","data":{"text/plain":"                             Name       Short Name  \\\n0                      IBM Maximo       IBM Maximo   \n1        Adobe Experience Manager              AEM   \n2  ServiceNow IT Asset Management  ServiceNow ITAM   \n\n                                         Description        Business Owner  \\\n0  A comprehensive enterprise asset management (E...    Strategic Planning   \n1  A digital asset management (DAM) system that h...    Strategic Planning   \n2  An IT asset management solution that integrate...  IT and Cybersecurity   \n\n              Technical Owner Number of Users Recovery Point Objective  \\\n0             Human Resources             415                      nan   \n1  Logistics and Supply Chain            3000                  60 mins   \n2        IT and Cybersecurity             176                      nan   \n\n  Recovery Time Objective Go Live Date End of Support Date  ...  \\\n0                     nan   2016-06-30          2023-06-30  ...   \n1          Zero data loss   2022-01-01          2024-11-30  ...   \n2                12 Hours   2011-04-01          2024-12-31  ...   \n\n  System Hosting Place User Community    Type of System Enterprise Level  \\\n0                Cloud           Both  System of Record               No   \n1           On Premise       Internal  System of Record              Yes   \n2           On Premise       Internal  System of Record              Yes   \n\n  Development Type Recommendation Cloud Migration Strategy Architecture Type  \\\n0             COTS        Migrate                   Retain         Mainframe   \n1             COTS         Retire                   Retire         Mainframe   \n2             COTS        Migrate               Replatform       Distributed   \n\n  Mobile Compliance Multi Language Support  \n0     Not Supported                    Yes  \n1     Not Supported                    Yes  \n2   Fully Supported                     No  \n\n[3 rows x 26 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Name</th>\n      <th>Short Name</th>\n      <th>Description</th>\n      <th>Business Owner</th>\n      <th>Technical Owner</th>\n      <th>Number of Users</th>\n      <th>Recovery Point Objective</th>\n      <th>Recovery Time Objective</th>\n      <th>Go Live Date</th>\n      <th>End of Support Date</th>\n      <th>...</th>\n      <th>System Hosting Place</th>\n      <th>User Community</th>\n      <th>Type of System</th>\n      <th>Enterprise Level</th>\n      <th>Development Type</th>\n      <th>Recommendation</th>\n      <th>Cloud Migration Strategy</th>\n      <th>Architecture Type</th>\n      <th>Mobile Compliance</th>\n      <th>Multi Language Support</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>IBM Maximo</td>\n      <td>IBM Maximo</td>\n      <td>A comprehensive enterprise asset management (E...</td>\n      <td>Strategic Planning</td>\n      <td>Human Resources</td>\n      <td>415</td>\n      <td>nan</td>\n      <td>nan</td>\n      <td>2016-06-30</td>\n      <td>2023-06-30</td>\n      <td>...</td>\n      <td>Cloud</td>\n      <td>Both</td>\n      <td>System of Record</td>\n      <td>No</td>\n      <td>COTS</td>\n      <td>Migrate</td>\n      <td>Retain</td>\n      <td>Mainframe</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>Adobe Experience Manager</td>\n      <td>AEM</td>\n      <td>A digital asset management (DAM) system that h...</td>\n      <td>Strategic Planning</td>\n      <td>Logistics and Supply Chain</td>\n      <td>3000</td>\n      <td>60 mins</td>\n      <td>Zero data loss</td>\n      <td>2022-01-01</td>\n      <td>2024-11-30</td>\n      <td>...</td>\n      <td>On Premise</td>\n      <td>Internal</td>\n      <td>System of Record</td>\n      <td>Yes</td>\n      <td>COTS</td>\n      <td>Retire</td>\n      <td>Retire</td>\n      <td>Mainframe</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>ServiceNow IT Asset Management</td>\n      <td>ServiceNow ITAM</td>\n      <td>An IT asset management solution that integrate...</td>\n      <td>IT and Cybersecurity</td>\n      <td>IT and Cybersecurity</td>\n      <td>176</td>\n      <td>nan</td>\n      <td>12 Hours</td>\n      <td>2011-04-01</td>\n      <td>2024-12-31</td>\n      <td>...</td>\n      <td>On Premise</td>\n      <td>Internal</td>\n      <td>System of Record</td>\n      <td>Yes</td>\n      <td>COTS</td>\n      <td>Migrate</td>\n      <td>Replatform</td>\n      <td>Distributed</td>\n      <td>Fully Supported</td>\n      <td>No</td>\n    </tr>\n  </tbody>\n</table>\n<p>3 rows × 26 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### 3. Add New Columns\n* Check if **new columns** formation is left w.r.t to the metadata\n* Add **newly created column** in the dataset","metadata":{}},{"cell_type":"code","source":"# Extract new columns data\nnew_columns = check_if_any_new_column_formation_is_required(df_uncleaned, metadata_json)\n\n# Create new columns\nfor new_column in new_columns:\n    new_column_name = new_column[0]\n    new_column_df = execute_create_new_column_prompt(new_column, extract_df_column_names_with_datatypes(df_uncleaned))\n    \n    # Concatenates column to the main dataset & remove duplicates column wise\n    if new_column_name in new_column_df.columns:\n        df_uncleaned = pd.concat([df_uncleaned, new_column_df], axis=1)\n        df_uncleaned = df_uncleaned.loc[:,~df_uncleaned.columns.duplicated()]\n        df_uncleaned = df_uncleaned.loc[:, ~df_uncleaned.columns.str.contains('^Unnamed')]","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:42.728881Z","iopub.execute_input":"2024-03-22T11:25:42.729319Z","iopub.status.idle":"2024-03-22T11:25:48.415663Z","shell.execute_reply.started":"2024-03-22T11:25:42.729285Z","shell.execute_reply":"2024-03-22T11:25:48.414026Z"},"trusted":true},"execution_count":29,"outputs":[]},{"cell_type":"code","source":"df_uncleaned.head(5)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:25:54.538988Z","iopub.execute_input":"2024-03-22T11:25:54.540441Z","iopub.status.idle":"2024-03-22T11:25:54.575933Z","shell.execute_reply.started":"2024-03-22T11:25:54.540382Z","shell.execute_reply":"2024-03-22T11:25:54.574703Z"},"trusted":true},"execution_count":30,"outputs":[{"execution_count":30,"output_type":"execute_result","data":{"text/plain":"                             Name       Short Name  \\\n0                      IBM Maximo       IBM Maximo   \n1        Adobe Experience Manager              AEM   \n2  ServiceNow IT Asset Management  ServiceNow ITAM   \n3               Sage Fixed Assets              SFA   \n4                         Flexera          Flexera   \n\n                                         Description        Business Owner  \\\n0  A comprehensive enterprise asset management (E...    Strategic Planning   \n1  A digital asset management (DAM) system that h...    Strategic Planning   \n2  An IT asset management solution that integrate...  IT and Cybersecurity   \n3  A fixed asset management software that offers ...      Fleet Management   \n4  A software asset management (SAM) tool that pr...   Marketing and Sales   \n\n              Technical Owner Number of Users Recovery Point Objective  \\\n0             Human Resources             415                      nan   \n1  Logistics and Supply Chain            3000                  60 mins   \n2        IT and Cybersecurity             176                      nan   \n3    Research and Development               4                      nan   \n4       Safety and Compliance              55                      nan   \n\n  Recovery Time Objective Go Live Date End of Support Date  ...  \\\n0                     nan   2016-06-30          2023-06-30  ...   \n1          Zero data loss   2022-01-01          2024-11-30  ...   \n2                12 Hours   2011-04-01          2024-12-31  ...   \n3                     nan          NaT                 NaT  ...   \n4                     nan          NaT          2020-04-30  ...   \n\n  User Community    Type of System Enterprise Level Development Type  \\\n0           Both  System of Record               No             COTS   \n1       Internal  System of Record              Yes             COTS   \n2       Internal  System of Record              Yes             COTS   \n3       Internal               NaN               No          Bespoke   \n4       Internal  System of Record              Yes          Bespoke   \n\n  Recommendation Cloud Migration Strategy Architecture Type Mobile Compliance  \\\n0        Migrate                   Retain         Mainframe     Not Supported   \n1         Retire                   Retire         Mainframe     Not Supported   \n2        Migrate               Replatform       Distributed   Fully Supported   \n3         Invest                   Retain       Stand-Alone     Not Supported   \n4       Tolerate               Repurchase           Unknown     Not Supported   \n\n  Multi Language Support Application Age  \n0                    Yes          2822.0  \n1                    Yes           811.0  \n2                     No          4739.0  \n3                    Yes             NaN  \n4                    Yes             NaN  \n\n[5 rows x 27 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Name</th>\n      <th>Short Name</th>\n      <th>Description</th>\n      <th>Business Owner</th>\n      <th>Technical Owner</th>\n      <th>Number of Users</th>\n      <th>Recovery Point Objective</th>\n      <th>Recovery Time Objective</th>\n      <th>Go Live Date</th>\n      <th>End of Support Date</th>\n      <th>...</th>\n      <th>User Community</th>\n      <th>Type of System</th>\n      <th>Enterprise Level</th>\n      <th>Development Type</th>\n      <th>Recommendation</th>\n      <th>Cloud Migration Strategy</th>\n      <th>Architecture Type</th>\n      <th>Mobile Compliance</th>\n      <th>Multi Language Support</th>\n      <th>Application Age</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>IBM Maximo</td>\n      <td>IBM Maximo</td>\n      <td>A comprehensive enterprise asset management (E...</td>\n      <td>Strategic Planning</td>\n      <td>Human Resources</td>\n      <td>415</td>\n      <td>nan</td>\n      <td>nan</td>\n      <td>2016-06-30</td>\n      <td>2023-06-30</td>\n      <td>...</td>\n      <td>Both</td>\n      <td>System of Record</td>\n      <td>No</td>\n      <td>COTS</td>\n      <td>Migrate</td>\n      <td>Retain</td>\n      <td>Mainframe</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n      <td>2822.0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>Adobe Experience Manager</td>\n      <td>AEM</td>\n      <td>A digital asset management (DAM) system that h...</td>\n      <td>Strategic Planning</td>\n      <td>Logistics and Supply Chain</td>\n      <td>3000</td>\n      <td>60 mins</td>\n      <td>Zero data loss</td>\n      <td>2022-01-01</td>\n      <td>2024-11-30</td>\n      <td>...</td>\n      <td>Internal</td>\n      <td>System of Record</td>\n      <td>Yes</td>\n      <td>COTS</td>\n      <td>Retire</td>\n      <td>Retire</td>\n      <td>Mainframe</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n      <td>811.0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>ServiceNow IT Asset Management</td>\n      <td>ServiceNow ITAM</td>\n      <td>An IT asset management solution that integrate...</td>\n      <td>IT and Cybersecurity</td>\n      <td>IT and Cybersecurity</td>\n      <td>176</td>\n      <td>nan</td>\n      <td>12 Hours</td>\n      <td>2011-04-01</td>\n      <td>2024-12-31</td>\n      <td>...</td>\n      <td>Internal</td>\n      <td>System of Record</td>\n      <td>Yes</td>\n      <td>COTS</td>\n      <td>Migrate</td>\n      <td>Replatform</td>\n      <td>Distributed</td>\n      <td>Fully Supported</td>\n      <td>No</td>\n      <td>4739.0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>Sage Fixed Assets</td>\n      <td>SFA</td>\n      <td>A fixed asset management software that offers ...</td>\n      <td>Fleet Management</td>\n      <td>Research and Development</td>\n      <td>4</td>\n      <td>nan</td>\n      <td>nan</td>\n      <td>NaT</td>\n      <td>NaT</td>\n      <td>...</td>\n      <td>Internal</td>\n      <td>NaN</td>\n      <td>No</td>\n      <td>Bespoke</td>\n      <td>Invest</td>\n      <td>Retain</td>\n      <td>Stand-Alone</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>Flexera</td>\n      <td>Flexera</td>\n      <td>A software asset management (SAM) tool that pr...</td>\n      <td>Marketing and Sales</td>\n      <td>Safety and Compliance</td>\n      <td>55</td>\n      <td>nan</td>\n      <td>nan</td>\n      <td>NaT</td>\n      <td>2020-04-30</td>\n      <td>...</td>\n      <td>Internal</td>\n      <td>System of Record</td>\n      <td>Yes</td>\n      <td>Bespoke</td>\n      <td>Tolerate</td>\n      <td>Repurchase</td>\n      <td>Unknown</td>\n      <td>Not Supported</td>\n      <td>Yes</td>\n      <td>NaN</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 27 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### 4. Generate Data card (Using Python Pandas)","metadata":{}},{"cell_type":"code","source":"# Scorecard Summary\nsummary = df_uncleaned.describe()\n\n# Data completeness\ncompleteness = df_uncleaned.notnull().sum()\ntotal_rows = len(df_uncleaned)\ncompleteness_percentage = (completeness / total_rows) * 100\n\n# Unique values\nunique_values = df_uncleaned.nunique()\n\n# Data types\ndata_types = df_uncleaned.dtypes\n\n# Creating data card dataframe\ncompleteness = pd.DataFrame({\n    'Completeness (%)': completeness_percentage,\n    'Unique Values': unique_values,\n    'Data Types': data_types\n})\n\n# Quality Issues\nquality_issues = pd.DataFrame({\n    'Data Type': df_uncleaned.dtypes,\n    'Unique Values': df_uncleaned.nunique(),\n    'Duplicates': df_uncleaned.duplicated().sum()\n})","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:26:02.927104Z","iopub.execute_input":"2024-03-22T11:26:02.927820Z","iopub.status.idle":"2024-03-22T11:26:02.978578Z","shell.execute_reply.started":"2024-03-22T11:26:02.927768Z","shell.execute_reply":"2024-03-22T11:26:02.977334Z"},"trusted":true},"execution_count":31,"outputs":[]},{"cell_type":"code","source":"summary","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:26:03.120368Z","iopub.execute_input":"2024-03-22T11:26:03.120869Z","iopub.status.idle":"2024-03-22T11:26:03.135925Z","shell.execute_reply.started":"2024-03-22T11:26:03.120832Z","shell.execute_reply":"2024-03-22T11:26:03.133965Z"},"trusted":true},"execution_count":32,"outputs":[{"execution_count":32,"output_type":"execute_result","data":{"text/plain":"                        Go Live Date            End of Support Date  \\\ncount                             57                             58   \nmean   2014-09-04 05:03:09.473684224  2023-10-29 02:04:08.275862016   \nmin              2004-01-01 00:00:00            2015-01-01 00:00:00   \n25%              2010-01-01 00:00:00            2023-01-12 06:00:00   \n50%              2015-01-01 00:00:00            2024-02-17 12:00:00   \n75%              2018-11-01 00:00:00            2024-11-30 00:00:00   \nmax              2023-07-28 00:00:00            2036-11-09 00:00:00   \nstd                              NaN                            NaN   \n\n       Application Age  \ncount        57.000000  \nmean       3486.789474  \nmin         238.000000  \n25%        1968.000000  \n50%        3368.000000  \n75%        5194.000000  \nmax        7386.000000  \nstd        1835.339826  ","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Go Live Date</th>\n      <th>End of Support Date</th>\n      <th>Application Age</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>57</td>\n      <td>58</td>\n      <td>57.000000</td>\n    </tr>\n    <tr>\n      <th>mean</th>\n      <td>2014-09-04 05:03:09.473684224</td>\n      <td>2023-10-29 02:04:08.275862016</td>\n      <td>3486.789474</td>\n    </tr>\n    <tr>\n      <th>min</th>\n      <td>2004-01-01 00:00:00</td>\n      <td>2015-01-01 00:00:00</td>\n      <td>238.000000</td>\n    </tr>\n    <tr>\n      <th>25%</th>\n      <td>2010-01-01 00:00:00</td>\n      <td>2023-01-12 06:00:00</td>\n      <td>1968.000000</td>\n    </tr>\n    <tr>\n      <th>50%</th>\n      <td>2015-01-01 00:00:00</td>\n      <td>2024-02-17 12:00:00</td>\n      <td>3368.000000</td>\n    </tr>\n    <tr>\n      <th>75%</th>\n      <td>2018-11-01 00:00:00</td>\n      <td>2024-11-30 00:00:00</td>\n      <td>5194.000000</td>\n    </tr>\n    <tr>\n      <th>max</th>\n      <td>2023-07-28 00:00:00</td>\n      <td>2036-11-09 00:00:00</td>\n      <td>7386.000000</td>\n    </tr>\n    <tr>\n      <th>std</th>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>1835.339826</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"quality_issues","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:26:04.718473Z","iopub.execute_input":"2024-03-22T11:26:04.719253Z","iopub.status.idle":"2024-03-22T11:26:04.735190Z","shell.execute_reply.started":"2024-03-22T11:26:04.719212Z","shell.execute_reply":"2024-03-22T11:26:04.733328Z"},"trusted":true},"execution_count":33,"outputs":[{"execution_count":33,"output_type":"execute_result","data":{"text/plain":"                               Data Type  Unique Values  Duplicates\nName                              object             98           0\nShort Name                        object             97           0\nDescription                       object            100           0\nBusiness Owner                  category             20           0\nTechnical Owner                 category             19           0\nNumber of Users                   object             49           0\nRecovery Point Objective          object             15           0\nRecovery Time Objective           object             14           0\nGo Live Date              datetime64[ns]             39           0\nEnd of Support Date       datetime64[ns]             43           0\nSLA Type                        category              4           0\nVendor                          category             30           0\nStatus                          category              3           0\nAGF Classification              category              4           0\nDR Type                         category              4           0\nRequired Availability           category              3           0\nSystem Hosting Place            category              3           0\nUser Community                  category              3           0\nType of System                  category              3           0\nEnterprise Level                category              2           0\nDevelopment Type                category              2           0\nRecommendation                  category              4           0\nCloud Migration Strategy        category              6           0\nArchitecture Type               category              7           0\nMobile Compliance               category              3           0\nMulti Language Support          category              2           0\nApplication Age                  float64             39           0","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Data Type</th>\n      <th>Unique Values</th>\n      <th>Duplicates</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>Name</th>\n      <td>object</td>\n      <td>98</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Short Name</th>\n      <td>object</td>\n      <td>97</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Description</th>\n      <td>object</td>\n      <td>100</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Business Owner</th>\n      <td>category</td>\n      <td>20</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Technical Owner</th>\n      <td>category</td>\n      <td>19</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Number of Users</th>\n      <td>object</td>\n      <td>49</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Recovery Point Objective</th>\n      <td>object</td>\n      <td>15</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Recovery Time Objective</th>\n      <td>object</td>\n      <td>14</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Go Live Date</th>\n      <td>datetime64[ns]</td>\n      <td>39</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>End of Support Date</th>\n      <td>datetime64[ns]</td>\n      <td>43</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>SLA Type</th>\n      <td>category</td>\n      <td>4</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Vendor</th>\n      <td>category</td>\n      <td>30</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Status</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>AGF Classification</th>\n      <td>category</td>\n      <td>4</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>DR Type</th>\n      <td>category</td>\n      <td>4</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Required Availability</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>System Hosting Place</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>User Community</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Type of System</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Enterprise Level</th>\n      <td>category</td>\n      <td>2</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Development Type</th>\n      <td>category</td>\n      <td>2</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Recommendation</th>\n      <td>category</td>\n      <td>4</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Cloud Migration Strategy</th>\n      <td>category</td>\n      <td>6</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Architecture Type</th>\n      <td>category</td>\n      <td>7</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Mobile Compliance</th>\n      <td>category</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Multi Language Support</th>\n      <td>category</td>\n      <td>2</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>Application Age</th>\n      <td>float64</td>\n      <td>39</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"completeness","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:26:04.950471Z","iopub.execute_input":"2024-03-22T11:26:04.952076Z","iopub.status.idle":"2024-03-22T11:26:04.973020Z","shell.execute_reply.started":"2024-03-22T11:26:04.952016Z","shell.execute_reply":"2024-03-22T11:26:04.971808Z"},"trusted":true},"execution_count":34,"outputs":[{"execution_count":34,"output_type":"execute_result","data":{"text/plain":"                          Completeness (%)  Unique Values      Data Types\nName                                 100.0             98          object\nShort Name                           100.0             97          object\nDescription                          100.0            100          object\nBusiness Owner                       100.0             20        category\nTechnical Owner                      100.0             19        category\nNumber of Users                      100.0             49          object\nRecovery Point Objective             100.0             15          object\nRecovery Time Objective              100.0             14          object\nGo Live Date                          57.0             39  datetime64[ns]\nEnd of Support Date                   58.0             43  datetime64[ns]\nSLA Type                              33.0              4        category\nVendor                               100.0             30        category\nStatus                                88.0              3        category\nAGF Classification                    93.0              4        category\nDR Type                               28.0              4        category\nRequired Availability                 45.0              3        category\nSystem Hosting Place                  62.0              3        category\nUser Community                        84.0              3        category\nType of System                        68.0              3        category\nEnterprise Level                      84.0              2        category\nDevelopment Type                     100.0              2        category\nRecommendation                       100.0              4        category\nCloud Migration Strategy             100.0              6        category\nArchitecture Type                    100.0              7        category\nMobile Compliance                     78.0              3        category\nMulti Language Support               100.0              2        category\nApplication Age                       57.0             39         float64","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Completeness (%)</th>\n      <th>Unique Values</th>\n      <th>Data Types</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>Name</th>\n      <td>100.0</td>\n      <td>98</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Short Name</th>\n      <td>100.0</td>\n      <td>97</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Description</th>\n      <td>100.0</td>\n      <td>100</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Business Owner</th>\n      <td>100.0</td>\n      <td>20</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Technical Owner</th>\n      <td>100.0</td>\n      <td>19</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Number of Users</th>\n      <td>100.0</td>\n      <td>49</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Recovery Point Objective</th>\n      <td>100.0</td>\n      <td>15</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Recovery Time Objective</th>\n      <td>100.0</td>\n      <td>14</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>Go Live Date</th>\n      <td>57.0</td>\n      <td>39</td>\n      <td>datetime64[ns]</td>\n    </tr>\n    <tr>\n      <th>End of Support Date</th>\n      <td>58.0</td>\n      <td>43</td>\n      <td>datetime64[ns]</td>\n    </tr>\n    <tr>\n      <th>SLA Type</th>\n      <td>33.0</td>\n      <td>4</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Vendor</th>\n      <td>100.0</td>\n      <td>30</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Status</th>\n      <td>88.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>AGF Classification</th>\n      <td>93.0</td>\n      <td>4</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>DR Type</th>\n      <td>28.0</td>\n      <td>4</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Required Availability</th>\n      <td>45.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>System Hosting Place</th>\n      <td>62.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>User Community</th>\n      <td>84.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Type of System</th>\n      <td>68.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Enterprise Level</th>\n      <td>84.0</td>\n      <td>2</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Development Type</th>\n      <td>100.0</td>\n      <td>2</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Recommendation</th>\n      <td>100.0</td>\n      <td>4</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Cloud Migration Strategy</th>\n      <td>100.0</td>\n      <td>6</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Architecture Type</th>\n      <td>100.0</td>\n      <td>7</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Mobile Compliance</th>\n      <td>78.0</td>\n      <td>3</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Multi Language Support</th>\n      <td>100.0</td>\n      <td>2</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>Application Age</th>\n      <td>57.0</td>\n      <td>39</td>\n      <td>float64</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"### 5. Ask RTA Bot (Talk with dataset)","metadata":{}},{"cell_type":"code","source":"ask_rta_bot('What is the User Community percentage w.r.t to its unique values', df_uncleaned)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:32:41.287610Z","iopub.execute_input":"2024-03-22T11:32:41.289631Z","iopub.status.idle":"2024-03-22T11:32:45.349778Z","shell.execute_reply.started":"2024-03-22T11:32:41.289545Z","shell.execute_reply":"2024-03-22T11:32:45.348615Z"},"trusted":true},"execution_count":38,"outputs":[{"execution_count":38,"output_type":"execute_result","data":{"text/plain":"  User Community  Percentage\n0       Internal   61.904762\n1           Both   21.428571\n2       External   16.666667","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>User Community</th>\n      <th>Percentage</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>Internal</td>\n      <td>61.904762</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>Both</td>\n      <td>21.428571</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>External</td>\n      <td>16.666667</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"ask_rta_bot('Check for data quality issues which should include missing values, unique values, duplicate values and their data types', df_uncleaned)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:32:48.158576Z","iopub.execute_input":"2024-03-22T11:32:48.159063Z","iopub.status.idle":"2024-03-22T11:32:51.394300Z","shell.execute_reply.started":"2024-03-22T11:32:48.159016Z","shell.execute_reply":"2024-03-22T11:32:51.393152Z"},"trusted":true},"execution_count":39,"outputs":[{"execution_count":39,"output_type":"execute_result","data":{"text/plain":"                  Unnamed: 0  Missing Values  Unique Values  Duplicate Values  \\\n0                       Name               0             98                 0   \n1                 Short Name               0             97                 0   \n2                Description               0            100                 0   \n3             Business Owner               0             20                 0   \n4            Technical Owner               0             19                 0   \n5            Number of Users               0             49                 0   \n6   Recovery Point Objective               0             15                 0   \n7    Recovery Time Objective               0             14                 0   \n8               Go Live Date              43             39                 0   \n9        End of Support Date              42             43                 0   \n10                  SLA Type              67              4                 0   \n11                    Vendor               0             30                 0   \n12                    Status              12              3                 0   \n13        AGF Classification               7              4                 0   \n14                   DR Type              72              4                 0   \n15     Required Availability              55              3                 0   \n16      System Hosting Place              38              3                 0   \n17            User Community              16              3                 0   \n18            Type of System              32              3                 0   \n19          Enterprise Level              16              2                 0   \n20          Development Type               0              2                 0   \n21            Recommendation               0              4                 0   \n22  Cloud Migration Strategy               0              6                 0   \n23         Architecture Type               0              7                 0   \n24         Mobile Compliance              22              3                 0   \n25    Multi Language Support               0              2                 0   \n26           Application Age              43             39                 0   \n\n        Data Types  \n0           object  \n1           object  \n2           object  \n3         category  \n4         category  \n5           object  \n6           object  \n7           object  \n8   datetime64[ns]  \n9   datetime64[ns]  \n10        category  \n11        category  \n12        category  \n13        category  \n14        category  \n15        category  \n16        category  \n17        category  \n18        category  \n19        category  \n20        category  \n21        category  \n22        category  \n23        category  \n24        category  \n25        category  \n26         float64  ","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Unnamed: 0</th>\n      <th>Missing Values</th>\n      <th>Unique Values</th>\n      <th>Duplicate Values</th>\n      <th>Data Types</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>Name</td>\n      <td>0</td>\n      <td>98</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>Short Name</td>\n      <td>0</td>\n      <td>97</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>Description</td>\n      <td>0</td>\n      <td>100</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>Business Owner</td>\n      <td>0</td>\n      <td>20</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>Technical Owner</td>\n      <td>0</td>\n      <td>19</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>Number of Users</td>\n      <td>0</td>\n      <td>49</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>Recovery Point Objective</td>\n      <td>0</td>\n      <td>15</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>Recovery Time Objective</td>\n      <td>0</td>\n      <td>14</td>\n      <td>0</td>\n      <td>object</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>Go Live Date</td>\n      <td>43</td>\n      <td>39</td>\n      <td>0</td>\n      <td>datetime64[ns]</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>End of Support Date</td>\n      <td>42</td>\n      <td>43</td>\n      <td>0</td>\n      <td>datetime64[ns]</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>SLA Type</td>\n      <td>67</td>\n      <td>4</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>Vendor</td>\n      <td>0</td>\n      <td>30</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>Status</td>\n      <td>12</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>AGF Classification</td>\n      <td>7</td>\n      <td>4</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>DR Type</td>\n      <td>72</td>\n      <td>4</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>Required Availability</td>\n      <td>55</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>System Hosting Place</td>\n      <td>38</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>User Community</td>\n      <td>16</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>18</th>\n      <td>Type of System</td>\n      <td>32</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>19</th>\n      <td>Enterprise Level</td>\n      <td>16</td>\n      <td>2</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>20</th>\n      <td>Development Type</td>\n      <td>0</td>\n      <td>2</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>21</th>\n      <td>Recommendation</td>\n      <td>0</td>\n      <td>4</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>22</th>\n      <td>Cloud Migration Strategy</td>\n      <td>0</td>\n      <td>6</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>23</th>\n      <td>Architecture Type</td>\n      <td>0</td>\n      <td>7</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>24</th>\n      <td>Mobile Compliance</td>\n      <td>22</td>\n      <td>3</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>25</th>\n      <td>Multi Language Support</td>\n      <td>0</td>\n      <td>2</td>\n      <td>0</td>\n      <td>category</td>\n    </tr>\n    <tr>\n      <th>26</th>\n      <td>Application Age</td>\n      <td>43</td>\n      <td>39</td>\n      <td>0</td>\n      <td>float64</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"ask_rta_bot('Plot Pie chart of Status', df_uncleaned)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:32:55.178708Z","iopub.execute_input":"2024-03-22T11:32:55.179868Z","iopub.status.idle":"2024-03-22T11:32:57.702497Z","shell.execute_reply.started":"2024-03-22T11:32:55.179821Z","shell.execute_reply":"2024-03-22T11:32:57.700839Z"},"trusted":true},"execution_count":40,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 800x800 with 1 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"code","source":"ask_rta_bot('Plot Bar graph of Status', df_uncleaned)","metadata":{"execution":{"iopub.status.busy":"2024-03-22T11:33:00.343995Z","iopub.execute_input":"2024-03-22T11:33:00.344581Z","iopub.status.idle":"2024-03-22T11:33:04.606356Z","shell.execute_reply.started":"2024-03-22T11:33:00.344546Z","shell.execute_reply":"2024-03-22T11:33:04.605150Z"},"trusted":true},"execution_count":41,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 1 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}