{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "markdown", "source": [ "## **Assignment #2: Classification, Regression, Clustering, Evaluation**" ], "metadata": { "id": "w84cR3AZIU0e" } }, { "cell_type": "markdown", "source": [ "`Version: April 2026`" ], "metadata": { "id": "sDxa7s952Ukh" } }, { "cell_type": "markdown", "source": [ "
" ], "metadata": { "id": "PnYmknSefeqx" } }, { "cell_type": "markdown", "source": [ "### **Overview**\n", "\n", "In this assignment, you'll level up your data science toolkit. While the first assignment focused on the data, on this one you will practice:\n", "\n", "- Classification models\n", "\n", "- Regression models\n", "\n", "- Feature Engineering\n", "\n", "- Evaluations\n", "\n", "You’ll go from raw data to insights by building a full modeling pipeline, enhancing your dataset, and training different models.\n", "\n", "This assignment will be completed individually." ], "metadata": { "id": "n7afdXdxIbLA" } }, { "cell_type": "markdown", "source": [ "### **Objectives**\n", "\n", "You’ll gain hands-on experience in:\n", "- Evaluation\n", "- Classification\n", "- Regression\n", "- Dataset preparation\n", "- Explore various data hubs\n", "- Engineering meaningful features\n", "- Communicating findings clearly - visually and verbally\n", "\n", "

" ], "metadata": { "id": "lJAPMumvIUyW" } }, { "cell_type": "markdown", "source": [ "### **Submission Guidelines**\n", "\n", "1. Please note that this assignmnet must be submitted alone.\n", "2. Submit the link to your HugingFace Model.\n", "\n", "Your HF model should include:\n", "- README file: explanations, visuals, insights, etc.\n", "- **Video**: Include the video of your presentation in the README file.\n", "- **Python Notebook**: upload a copy of this notebook, with all of your coding work. Do not submit a Colab link; include the `.ipynb` file in the HF model.\n", "- **ML Models:** Upload your models.\n", "\n", "Note: Students may be randomly chosen to present their work in a quick online session with the T.A., typically lasting ±10 minutes. Similar to Peer Review.\n", "\n", "

\n", "\n" ], "metadata": { "id": "MwRmaJBiIjMR" } }, { "cell_type": "markdown", "source": [ "### **Evaluation Criteria**\n", "\n", "* **Data Handling & EDA (20%)**\n", " Thoughtful and thorough data cleaning; handling of missing values, outliers, duplicates, and more; well-chosen visualizations; clear statistical summaries; use of EDA to guide modeling choices.\n", "\n", "* **Feature Engineering (20%)**\n", " Creative and effective feature creation, transformation, encoding, selection, scaling, and more; integration of clustering results as features; clear explanation of feature choices and their impact.\n", "\n", "* **Model Training (20%)**\n", " Appropriate selection of models; correct train/test split; reproducible code; logical modeling workflow with a solid baseline and improvements post-feature engineering. An iterative process.\n", "\n", "* **Evaluation & Interpretation (20%)**\n", " Use of relevant evaluation metrics; structured model comparison; use of feature importance or visualizations to interpret results; clear discussion of what the model learned and how it performed.\n", "\n", "* **Presentation (20%)**\n", " 4–6 minute video with clear delivery; structured narrative; visuals that support the explanation; confident, professional communication of findings and lessons.\n", "\n", "* **Bonus (up to +10%)**\n", " Extra work such as trying data science tools, creative visualizations, advanced hyper param tuning, interactive dashboards, and deeper business/domain insights.\n", "\n", "* **Late Submission (-10% per day)**\n", " Assignments submitted after the deadline will receive a 10% penalty per day.\n", "\n", "

" ], "metadata": { "id": "hD9SZmagIjOV" } }, { "cell_type": "markdown", "source": [ "### **Additional Guidelines**\n", "\n", "- The first thing you should do is download a copy of this notebook to your drive.\n", "- Keep your dataset size manageable. If the dataset is too large, you can sample a subset.\n", "- Run on Colab (CPU is fine). Colab free is enough. No GPU needed.\n", "- You may use any Python package (scikit-learn, xgboost, lightgbm, catboost, etc.).\n", "- No SHAP required. Use `feature_importances`, and similar tools.\n", "- Make sure your results are reproducible (set **seeds** where needed).\n", "- Be thoughtful with your cluster features — only use them if they help!\n", "- Your presentation should tell a story; what worked, what didn’t, and why.\n", "- Be creative, but also rigorous." ], "metadata": { "id": "h3vpVHSxIUwI" } }, { "cell_type": "markdown", "source": [ "### Assignment High-level Flow" ], "metadata": { "id": "7lTH1B5b5c12" } }, { "cell_type": "markdown", "source": [ "![image.png](data:image/png;base64,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)" ], "metadata": { "id": "EK9fe2XygjgM" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "6kUonEv8Ipkp" } }, { "cell_type": "markdown", "source": [ "imports" ], "metadata": { "id": "acyYQrhPdEhB" } }, { "cell_type": "code", "source": [ "import os\n", "import random\n", "\n", "import numpy as np\n", "import pandas as pd\n", "\n", "import seaborn\n", "import matplotlib.pyplot as plt\n", "\n", "# " ], "metadata": { "id": "H9ZazAMOc5jC" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "Set Seeds" ], "metadata": { "id": "TPaWKBWmdGNF" } }, { "cell_type": "code", "source": [ "SEED = 42\n", "\n", "random.seed(SEED)\n", "np.random.seed(SEED)\n", "os.environ['PYTHONHASHSEED'] = str(SEED)\n", "\n", "# " ], "metadata": { "id": "zBaCUY21dHeF" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "For Jupyter Notebooks" ], "metadata": { "id": "INAizD1WeZcf" } }, { "cell_type": "code", "source": [ "%matplotlib inline\n", "%config InlineBackend.figure_format = 'retina'" ], "metadata": { "id": "G0hg5eohd4s-" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "Warnings" ], "metadata": { "id": "UwSXkPGvecLK" } }, { "cell_type": "code", "source": [ "import warnings\n", "warnings.filterwarnings('ignore')" ], "metadata": { "id": "Nk9C78G3d7vp" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "

\n", "\n", "---\n", "\n", "

" ], "metadata": { "id": "evFmlLbzdgBj" } }, { "cell_type": "markdown", "source": [ "# **Part 1: Select a Regression Dataset**\n", "\n", "1. Choose a numeric & categorical tabular dataset. If you prefer, you may use open-source datasets; [Hugginface](https://huggingface.co/datasets?task_categories=task_categories:tabular-classification&sort=trending), [Kaggle](https://www.kaggle.com/datasets?tags=13302-Classification&minUsabilityRating=8.00+or+higher), etc.\n", "\n", "2. Avoid choosing a \"basic\"/\"small\" dataset.\n", " - 10K rows and more.\n", " - 15 features and more.\n", " - Mix of Numeric & Categorial features are a must.\n", "\n", "3. The Label (target variable) is numeric.\n", "\n", "4. Please submit your dataset [here](https://forms.gle/zS8aZbBzuBV2z7wZ7), to share it with the class so everyone can see.\n", "And make sure your chosen dataset is unique using this [link](https://docs.google.com/spreadsheets/d/1M8uojrzhSyVnOlSAJpzCKxrhWdzPR77k4x8Kxvr8VDk/edit?usp=sharing).\n", "\n", " *Note: Due to their popularity, the following are datasets you may not choose.*\n", " > - Iris dataset\n", " > - Wine dataset\n", " > - Titanic dataset\n", " > - Boston Housing dataset\n", " > - ImageNet, Cifar, CelebFaces, IMDB\n", "\n", "5. Choose a dataset with a combination of numeric and textual values. This way you would have enough information to work on.\n", "\n", "6. Briefly describe your chosen dataset (source, size, features) and the question you want to answer." ], "metadata": { "id": "a_vsO0Q1IOMT" } }, { "cell_type": "markdown", "source": [ "### Dataset: DataCo Smart Supply Chain\n", "\n", "**Source:** Kaggle `DataCoSupplyChainDataset.csv` \n", "**Size:** 180,519 rows, 53 columns\n", "\n", "The dataset covers global e-commerce supply chain operations, tracking orders from purchase through delivery. It contains a mix of numeric and categorical features including order financials, shipping details, customer segments, and geographic data, along with real-world data quality issues like missing values and encoding inconsistencies.\n", "\n", "**Key features:**\n", "- `Benefit per order`: profit or loss per shipment in USD\n", "- `Days for shipping (real)` vs `Days for shipment (scheduled)`: actual vs. planned delivery time.\n", "- `Late_delivery_risk`: whether the order was flagged as a late delivery risk (0/1)\n", "- `Shipping Mode`, `Customer Segment`, `Order Region`, `Category Name`, `Type` (payment method)\n", "\n", "**The question I want to answer:**\n", "\n", "Can we predict, at the time an order is placed, whether it will arrive late and how profitable it will be?\n", "\n", "To answer this I will build two models:\n", "- A **regression model** to predict `Benefit per order`\n", "- A **classification model** to predict `Late_delivery_risk`\n", "\n", "Knowing in advance which orders are likely to be late or unprofitable allows a business to intervene early, re-route shipments, or adjust pricing strategies." ], "metadata": { "id": "XCOMMBwIMOWU" } }, { "cell_type": "markdown", "source": [], "metadata": { "id": "GFaHgQJyKwa5" } }, { "cell_type": "code", "source": [ "# Load dataset directly from HuggingFace — no Kaggle token needed\n", "import pandas as pd\n", "\n", "url = 'https://huggingface.co/idoyaaran/dataco-supply-chain-model/resolve/main/DataCoSupplyChainDataset.csv'\n", "df = pd.read_csv(url, encoding='cp1252', on_bad_lines='skip')\n", "print(df.shape)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "lzLvpoRNhv3j", "outputId": "b9376351-9c66-4459-de8e-ff22cdc94261" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180519, 53)\n" ] } ] }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "4t2QNyE6IPKS" } }, { "cell_type": "markdown", "source": [ "# **Part 2: Exploratory Data Analysis (EDA)**\n", "\n", "Use your EDA to tell the story of your data - highlight interesting patterns, anomalies, or relationships that lead you toward your classification goal. Ask interesting questions, and answer them.\n", "\n", "\n", "1. **Data Cleaning** : Check for missing values, duplicate entries, scaling/normalize issues, parsing dates, fixing typos, or any inconsistencies. Document how you address them.\n", "2. **Outlier Detection & Handling**: Identify outliers and decide whether to keep or remove them, providing a short justification.\n", "2. **Descriptive Statistics**: Summarize the data (e.g., mean, median, correlations) to reveal patterns.\n", "4. **Visualizations**: Use a set of plots (e.g., histograms, scatter plots, box plots) to illustrate **key insights.** Label charts, axes, and legends clearly.\n", "\n", "Tip: not necessarily in this order." ], "metadata": { "id": "6eLmNWJJIPS0" } }, { "cell_type": "code", "source": [ "df.info()" ], "metadata": { "id": "fUEbi2wOhssS", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "753b2ad2-210a-45ba-aff4-de0b878b58af" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\n", "RangeIndex: 180519 entries, 0 to 180518\n", "Data columns (total 53 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 Type 180519 non-null object \n", " 1 Days for shipping (real) 180519 non-null int64 \n", " 2 Days for shipment (scheduled) 180519 non-null int64 \n", " 3 Benefit per order 180519 non-null float64\n", " 4 Sales per customer 180519 non-null float64\n", " 5 Delivery Status 180519 non-null object \n", " 6 Late_delivery_risk 180519 non-null int64 \n", " 7 Category Id 180519 non-null int64 \n", " 8 Category Name 180519 non-null object \n", " 9 Customer City 180519 non-null object \n", " 10 Customer Country 180519 non-null object \n", " 11 Customer Email 180519 non-null object \n", " 12 Customer Fname 180519 non-null object \n", " 13 Customer Id 180519 non-null int64 \n", " 14 Customer Lname 180511 non-null object \n", " 15 Customer Password 180519 non-null object \n", " 16 Customer Segment 180519 non-null object \n", " 17 Customer State 180519 non-null object \n", " 18 Customer Street 180519 non-null object \n", " 19 Customer Zipcode 180516 non-null float64\n", " 20 Department Id 180519 non-null int64 \n", " 21 Department Name 180519 non-null object \n", " 22 Latitude 180519 non-null float64\n", " 23 Longitude 180519 non-null float64\n", " 24 Market 180519 non-null object \n", " 25 Order City 180519 non-null object \n", " 26 Order Country 180519 non-null object \n", " 27 Order Customer Id 180519 non-null int64 \n", " 28 order date (DateOrders) 180519 non-null object \n", " 29 Order Id 180519 non-null int64 \n", " 30 Order Item Cardprod Id 180519 non-null int64 \n", " 31 Order Item Discount 180519 non-null float64\n", " 32 Order Item Discount Rate 180519 non-null float64\n", " 33 Order Item Id 180519 non-null int64 \n", " 34 Order Item Product Price 180519 non-null float64\n", " 35 Order Item Profit Ratio 180519 non-null float64\n", " 36 Order Item Quantity 180519 non-null int64 \n", " 37 Sales 180519 non-null float64\n", " 38 Order Item Total 180519 non-null float64\n", " 39 Order Profit Per Order 180519 non-null float64\n", " 40 Order Region 180519 non-null object \n", " 41 Order State 180519 non-null object \n", " 42 Order Status 180519 non-null object \n", " 43 Order Zipcode 24840 non-null float64\n", " 44 Product Card Id 180519 non-null int64 \n", " 45 Product Category Id 180519 non-null int64 \n", " 46 Product Description 0 non-null float64\n", " 47 Product Image 180519 non-null object \n", " 48 Product Name 180519 non-null object \n", " 49 Product Price 180519 non-null float64\n", " 50 Product Status 180519 non-null int64 \n", " 51 shipping date (DateOrders) 180519 non-null object \n", " 52 Shipping Mode 180519 non-null object \n", "dtypes: float64(15), int64(14), object(24)\n", "memory usage: 73.0+ MB\n" ] } ] }, { "cell_type": "code", "source": [ "df.describe()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 380 }, "id": "qPNzI2KkO0JM", "outputId": "8dd70132-e43f-4ab1-c93a-18cd3983d61c" }, "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " Days for shipping (real) Days for shipment (scheduled) \\\n", "count 180519.000000 180519.000000 \n", "mean 3.497654 2.931847 \n", "std 1.623722 1.374449 \n", "min 0.000000 0.000000 \n", "25% 2.000000 2.000000 \n", "50% 3.000000 4.000000 \n", "75% 5.000000 4.000000 \n", "max 6.000000 4.000000 \n", "\n", " Benefit per order Sales per customer Late_delivery_risk \\\n", "count 180519.000000 180519.000000 180519.000000 \n", "mean 21.974989 183.107609 0.548291 \n", "std 104.433526 120.043670 0.497664 \n", "min -4274.979980 7.490000 0.000000 \n", "25% 7.000000 104.379997 0.000000 \n", "50% 31.520000 163.990005 1.000000 \n", "75% 64.800003 247.399994 1.000000 \n", "max 911.799988 1939.989990 1.000000 \n", "\n", " Category Id Customer Id Customer Zipcode Department Id \\\n", "count 180519.000000 180519.000000 180516.000000 180519.000000 \n", "mean 31.851451 6691.379495 35921.126914 5.443460 \n", "std 15.640064 4162.918106 37542.461122 1.629246 \n", "min 2.000000 1.000000 603.000000 2.000000 \n", "25% 18.000000 3258.500000 725.000000 4.000000 \n", "50% 29.000000 6457.000000 19380.000000 5.000000 \n", "75% 45.000000 9779.000000 78207.000000 7.000000 \n", "max 76.000000 20757.000000 99205.000000 12.000000 \n", "\n", " Latitude ... Order Item Quantity Sales \\\n", "count 180519.000000 ... 180519.000000 180519.000000 \n", "mean 29.719955 ... 2.127638 203.772096 \n", "std 9.813646 ... 1.453451 132.273077 \n", "min -33.937553 ... 1.000000 9.990000 \n", "25% 18.265432 ... 1.000000 119.980003 \n", "50% 33.144863 ... 1.000000 199.919998 \n", "75% 39.279617 ... 3.000000 299.950012 \n", "max 48.781933 ... 5.000000 1999.989990 \n", "\n", " Order Item Total Order Profit Per Order Order Zipcode \\\n", "count 180519.000000 180519.000000 24840.000000 \n", "mean 183.107609 21.974989 55426.132327 \n", "std 120.043670 104.433526 31919.279101 \n", "min 7.490000 -4274.979980 1040.000000 \n", "25% 104.379997 7.000000 23464.000000 \n", "50% 163.990005 31.520000 59405.000000 \n", "75% 247.399994 64.800003 90008.000000 \n", "max 1939.989990 911.799988 99301.000000 \n", "\n", " Product Card Id Product Category Id Product Description \\\n", "count 180519.000000 180519.000000 0.0 \n", "mean 692.509764 31.851451 NaN \n", "std 336.446807 15.640064 NaN \n", "min 19.000000 2.000000 NaN \n", "25% 403.000000 18.000000 NaN \n", "50% 627.000000 29.000000 NaN \n", "75% 1004.000000 45.000000 NaN \n", "max 1363.000000 76.000000 NaN \n", "\n", " Product Price Product Status \n", "count 180519.000000 180519.0 \n", "mean 141.232550 0.0 \n", "std 139.732492 0.0 \n", "min 9.990000 0.0 \n", "25% 50.000000 0.0 \n", "50% 59.990002 0.0 \n", "75% 199.990005 0.0 \n", "max 1999.989990 0.0 \n", "\n", "[8 rows x 29 columns]" ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Days for shipping (real)Days for shipment (scheduled)Benefit per orderSales per customerLate_delivery_riskCategory IdCustomer IdCustomer ZipcodeDepartment IdLatitude...Order Item QuantitySalesOrder Item TotalOrder Profit Per OrderOrder ZipcodeProduct Card IdProduct Category IdProduct DescriptionProduct PriceProduct Status
count180519.000000180519.000000180519.000000180519.000000180519.000000180519.000000180519.000000180516.000000180519.000000180519.000000...180519.000000180519.000000180519.000000180519.00000024840.000000180519.000000180519.0000000.0180519.000000180519.0
mean3.4976542.93184721.974989183.1076090.54829131.8514516691.37949535921.1269145.44346029.719955...2.127638203.772096183.10760921.97498955426.132327692.50976431.851451NaN141.2325500.0
std1.6237221.374449104.433526120.0436700.49766415.6400644162.91810637542.4611221.6292469.813646...1.453451132.273077120.043670104.43352631919.279101336.44680715.640064NaN139.7324920.0
min0.0000000.000000-4274.9799807.4900000.0000002.0000001.000000603.0000002.000000-33.937553...1.0000009.9900007.490000-4274.9799801040.00000019.0000002.000000NaN9.9900000.0
25%2.0000002.0000007.000000104.3799970.00000018.0000003258.500000725.0000004.00000018.265432...1.000000119.980003104.3799977.00000023464.000000403.00000018.000000NaN50.0000000.0
50%3.0000004.00000031.520000163.9900051.00000029.0000006457.00000019380.0000005.00000033.144863...1.000000199.919998163.99000531.52000059405.000000627.00000029.000000NaN59.9900020.0
75%5.0000004.00000064.800003247.3999941.00000045.0000009779.00000078207.0000007.00000039.279617...3.000000299.950012247.39999464.80000390008.0000001004.00000045.000000NaN199.9900050.0
max6.0000004.000000911.7999881939.9899901.00000076.00000020757.00000099205.00000012.00000048.781933...5.0000001999.9899901939.989990911.79998899301.0000001363.00000076.000000NaN1999.9899900.0
\n", "

8 rows × 29 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe" } }, "metadata": {}, "execution_count": 8 } ] }, { "cell_type": "code", "source": [ "# check missing values\n", "df.isnull().sum()[df.isnull().sum() > 0]" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 207 }, "id": "pGGJJwRkO1-M", "outputId": "19dda92d-c5d4-4748-9b39-3d5fe5fee17e" }, "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "Customer Lname 8\n", "Customer Zipcode 3\n", "Order Zipcode 155679\n", "Product Description 180519\n", "dtype: int64" ], "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
0
Customer Lname8
Customer Zipcode3
Order Zipcode155679
Product Description180519
\n", "

" ] }, "metadata": {}, "execution_count": 9 } ] }, { "cell_type": "code", "source": [ "# columns to drop\n", "cols_to_drop = [\n", " # PII\n", " 'Customer Email', 'Customer Password', 'Customer Fname',\n", " 'Customer Lname', 'Customer Street',\n", " # empty / high missing\n", " 'Product Description', 'Order Zipcode',\n", " # useless for modeling\n", " 'Product Image',\n", " # redundant IDs\n", " 'Customer Id', 'Order Customer Id', 'Order Id', 'Order Item Id',\n", " 'Order Item Cardprod Id', 'Product Card Id',\n", " 'Category Id', 'Department Id', 'Product Category Id'\n", "]\n", "\n", "df.drop(columns=cols_to_drop, inplace=True)\n", "\n", "# fix date columns\n", "df['order date (DateOrders)'] = pd.to_datetime(df['order date (DateOrders)'])\n", "df['shipping date (DateOrders)'] = pd.to_datetime(df['shipping date (DateOrders)'])\n", "\n", "print(df.shape)\n", "print(f\"Missing values remaining:\\n{df.isnull().sum()[df.isnull().sum() > 0]}\")" ], "metadata": { "id": "Qe5k-9DLK07M", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "5b63a0b8-f295-45a5-ac0c-4208a96a51bc" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180519, 36)\n", "Missing values remaining:\n", "Customer Zipcode 3\n", "dtype: int64\n" ] } ] }, { "cell_type": "code", "source": [ "df.dropna(subset=['Customer Zipcode'], inplace=True)\n", "print(df.shape)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "w2g8bjSyPmdD", "outputId": "30aab606-0879-4b0f-a009-c7b407483630" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180516, 36)\n" ] } ] }, { "cell_type": "code", "source": [ "df.describe()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 453 }, "id": "mOosyBTZQkDI", "outputId": "3da8c5d5-939c-4daa-8f46-8545cce46709" }, "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " Days for shipping (real) Days for shipment (scheduled) \\\n", "count 180516.000000 180516.000000 \n", "mean 3.497657 2.931845 \n", "min 0.000000 0.000000 \n", "25% 2.000000 2.000000 \n", "50% 3.000000 4.000000 \n", "75% 5.000000 4.000000 \n", "max 6.000000 4.000000 \n", "std 1.623731 1.374449 \n", "\n", " Benefit per order Sales per customer Late_delivery_risk \\\n", "count 180516.000000 180516.000000 180516.000000 \n", "mean 21.974593 183.107723 0.548295 \n", "min -4274.979980 7.490000 0.000000 \n", "25% 7.000000 104.379997 0.000000 \n", "50% 31.520000 163.990005 1.000000 \n", "75% 64.800003 247.399994 1.000000 \n", "max 911.799988 1939.989990 1.000000 \n", "std 104.434300 120.043872 0.497664 \n", "\n", " Customer Zipcode Latitude Longitude \\\n", "count 180516.000000 180516.000000 180516.000000 \n", "mean 35921.126914 29.719830 -84.915074 \n", "min 603.000000 -33.937553 -158.025986 \n", "25% 725.000000 18.265432 -98.446312 \n", "50% 19380.000000 33.144863 -76.840759 \n", "75% 78207.000000 39.279617 -66.370583 \n", "max 99205.000000 48.781933 115.263077 \n", "std 37542.461122 9.813676 21.432911 \n", "\n", " order date (DateOrders) Order Item Discount \\\n", "count 180516 180516.000000 \n", "mean 2016-06-12 17:34:40.783421184 20.664413 \n", "min 2015-01-01 00:00:00 0.000000 \n", "25% 2015-09-21 13:28:00 5.400000 \n", "50% 2016-06-11 13:06:00 14.000000 \n", "75% 2017-03-01 08:42:00 29.990000 \n", "max 2018-01-31 23:38:00 500.000000 \n", "std NaN 21.800633 \n", "\n", " Order Item Discount Rate Order Item Product Price \\\n", "count 180516.000000 180516.000000 \n", "mean 0.101667 141.231298 \n", "min 0.000000 9.990000 \n", "25% 0.040000 50.000000 \n", "50% 0.100000 59.990002 \n", "75% 0.160000 199.990005 \n", "max 0.250000 1999.989990 \n", "std 0.070415 139.732278 \n", "\n", " Order Item Profit Ratio Order Item Quantity Sales \\\n", "count 180516.000000 180516.000000 180516.000000 \n", "mean 0.120644 2.127656 203.771883 \n", "min -2.750000 1.000000 9.990000 \n", "25% 0.080000 1.000000 119.980003 \n", "50% 0.270000 1.000000 199.919998 \n", "75% 0.360000 3.000000 299.950012 \n", "max 0.500000 5.000000 1999.989990 \n", "std 0.466799 1.453456 132.273070 \n", "\n", " Order Item Total Order Profit Per Order Product Price \\\n", "count 180516.000000 180516.000000 180516.000000 \n", "mean 183.107723 21.974593 141.231298 \n", "min 7.490000 -4274.979980 9.990000 \n", "25% 104.379997 7.000000 50.000000 \n", "50% 163.990005 31.520000 59.990002 \n", "75% 247.399994 64.800003 199.990005 \n", "max 1939.989990 911.799988 1999.989990 \n", "std 120.043872 104.434300 139.732278 \n", "\n", " Product Status shipping date (DateOrders) \n", "count 180516.0 180516 \n", "mean 0.0 2016-06-16 05:32:59.553612544 \n", "min 0.0 2015-01-03 00:00:00 \n", "25% 0.0 2015-09-25 06:40:30 \n", "50% 0.0 2016-06-15 07:29:00 \n", "75% 0.0 2017-03-04 20:41:45 \n", "max 0.0 2018-02-06 22:14:00 \n", "std 0.0 NaN " ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Days for shipping (real)Days for shipment (scheduled)Benefit per orderSales per customerLate_delivery_riskCustomer ZipcodeLatitudeLongitudeorder date (DateOrders)Order Item DiscountOrder Item Discount RateOrder Item Product PriceOrder Item Profit RatioOrder Item QuantitySalesOrder Item TotalOrder Profit Per OrderProduct PriceProduct Statusshipping date (DateOrders)
count180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.000000180516.0180516
mean3.4976572.93184521.974593183.1077230.54829535921.12691429.719830-84.9150742016-06-12 17:34:40.78342118420.6644130.101667141.2312980.1206442.127656203.771883183.10772321.974593141.2312980.02016-06-16 05:32:59.553612544
min0.0000000.000000-4274.9799807.4900000.000000603.000000-33.937553-158.0259862015-01-01 00:00:000.0000000.0000009.990000-2.7500001.0000009.9900007.490000-4274.9799809.9900000.02015-01-03 00:00:00
25%2.0000002.0000007.000000104.3799970.000000725.00000018.265432-98.4463122015-09-21 13:28:005.4000000.04000050.0000000.0800001.000000119.980003104.3799977.00000050.0000000.02015-09-25 06:40:30
50%3.0000004.00000031.520000163.9900051.00000019380.00000033.144863-76.8407592016-06-11 13:06:0014.0000000.10000059.9900020.2700001.000000199.919998163.99000531.52000059.9900020.02016-06-15 07:29:00
75%5.0000004.00000064.800003247.3999941.00000078207.00000039.279617-66.3705832017-03-01 08:42:0029.9900000.160000199.9900050.3600003.000000299.950012247.39999464.800003199.9900050.02017-03-04 20:41:45
max6.0000004.000000911.7999881939.9899901.00000099205.00000048.781933115.2630772018-01-31 23:38:00500.0000000.2500001999.9899900.5000005.0000001999.9899901939.989990911.7999881999.9899900.02018-02-06 22:14:00
std1.6237311.374449104.434300120.0438720.49766437542.4611229.81367621.432911NaN21.8006330.070415139.7322780.4667991.453456132.273070120.043872104.434300139.7322780.0NaN
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "summary": "{\n \"name\": \"df\",\n \"rows\": 8,\n \"fields\": [\n {\n \"column\": \"Days for shipping (real)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63820.977093312926,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 3.497656717410091,\n 5.0,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Days for shipment (scheduled)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63821.1192652163,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 6,\n \"samples\": [\n 180516.0,\n 2.931845376587117,\n 1.3744485815196124\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Benefit per order\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 64000.087765597185,\n \"min\": -4274.97998,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 21.9745933548914,\n 64.80000305,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Sales per customer\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63685.47935607165,\n \"min\": 7.489999771,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 183.10772341633054,\n 247.3999939,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Late_delivery_risk\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63821.83950593046,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 5,\n \"samples\": [\n 0.5482948879877684,\n 0.4976635166459778,\n 0.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Customer Zipcode\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 61028.06827549364,\n \"min\": 603.0,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 35921.12691395776,\n 78207.0,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Latitude\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63814.72184024707,\n \"min\": -33.93755341,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 29.719830269997498,\n 39.27961731,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Longitude\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63839.670017437296,\n \"min\": -158.0259857,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n -84.915073995796,\n -66.37058258,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"order date (DateOrders)\",\n \"properties\": {\n \"dtype\": \"date\",\n \"min\": \"1970-01-01 00:00:00.000180516\",\n \"max\": \"2018-01-31 23:38:00\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"180516\",\n \"2016-06-12 17:34:40.783421184\",\n \"2017-03-01 08:42:00\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Discount\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63792.376826167,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 20.6644133627598,\n 29.98999977,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Discount Rate\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63822.0073857306,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 0.10166749811797844,\n 0.159999996,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Product Price\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63694.17546773381,\n \"min\": 9.989999771,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 141.23129778284536,\n 199.9900055,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Profit Ratio\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63822.091976819305,\n \"min\": -2.75,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 0.12064404258940481,\n 0.360000014,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Quantity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63821.30741378561,\n \"min\": 1.0,\n \"max\": 180516.0,\n \"num_unique_values\": 6,\n \"samples\": [\n 180516.0,\n 2.1276562742360787,\n 1.4534562891340808\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Sales\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63675.5502117577,\n \"min\": 9.989999771,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 203.77188328662749,\n 299.9500122,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Item Total\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63685.47935607165,\n \"min\": 7.489999771,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 183.10772341633054,\n 247.3999939,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Order Profit Per Order\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 64000.087765597185,\n \"min\": -4274.97998,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 21.9745933548914,\n 64.80000305,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Product Price\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63694.17546773381,\n \"min\": 9.989999771,\n \"max\": 180516.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 141.23129778284536,\n 199.9900055,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Product Status\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 63822.043856335404,\n \"min\": 0.0,\n \"max\": 180516.0,\n \"num_unique_values\": 2,\n \"samples\": [\n 0.0,\n 180516.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"shipping date (DateOrders)\",\n \"properties\": {\n \"dtype\": \"date\",\n \"min\": \"1970-01-01 00:00:00.000180516\",\n \"max\": \"2018-02-06 22:14:00\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"180516\",\n \"2016-06-16 05:32:59.553612544\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 12 } ] }, { "cell_type": "code", "source": [ "df.drop(columns=['Product Status'], inplace=True)" ], "metadata": { "id": "xtxF-Aj2RFIG" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "source": [ "# verify they're identical\n", "print((df['Benefit per order'] == df['Order Profit Per Order']).all())\n", "print((df['Sales per customer'] == df['Order Item Total']).all())\n", "print((df['Order Item Product Price'] == df['Product Price']).all())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "uD9QZ1BdRHdR", "outputId": "ddd09b71-73af-4f49-94c3-29f5e8248b8b" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "True\n", "True\n", "True\n" ] } ] }, { "cell_type": "code", "source": [ "df.drop(columns=['Order Profit Per Order', 'Order Item Total', 'Product Price'], inplace=True)\n", "print(df.shape)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "vu6MFEciRJrx", "outputId": "34539927-7121-4f7d-a8f4-c01fa85c524f" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180516, 32)\n" ] } ] }, { "cell_type": "code", "source": [ "print(df['Order Item Discount'].describe())\n", "print(df[df['Order Item Discount'] > 200])" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "v5OzfG1TRSNS", "outputId": "4c866510-9982-4b33-b5a8-1561b537784e" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "count 180516.000000\n", "mean 20.664413\n", "std 21.800633\n", "min 0.000000\n", "25% 5.400000\n", "50% 14.000000\n", "75% 29.990000\n", "max 500.000000\n", "Name: Order Item Discount, dtype: float64\n", " Type Days for shipping (real) Days for shipment (scheduled) \\\n", "172 DEBIT 5 2 \n", "1353 PAYMENT 3 4 \n", "1370 PAYMENT 2 4 \n", "1375 PAYMENT 5 4 \n", "1594 DEBIT 6 2 \n", "... ... ... ... \n", "170177 TRANSFER 2 4 \n", "170308 TRANSFER 4 4 \n", "170841 TRANSFER 3 4 \n", "173423 CASH 4 4 \n", "173932 CASH 6 4 \n", "\n", " Benefit per order Sales per customer Delivery Status \\\n", "172 90.000000 1200.0 Late delivery \n", "1353 82.410004 1230.0 Advance shipping \n", "1370 599.250000 1275.0 Advance shipping \n", "1375 410.850006 1245.0 Late delivery \n", "1594 281.250000 1125.0 Late delivery \n", "... ... ... ... \n", "170177 375.600006 1200.0 Advance shipping \n", "170308 466.880005 1245.0 Shipping canceled \n", "170841 155.630005 1245.0 Advance shipping \n", "173423 127.500000 1275.0 Shipping on time \n", "173932 -320.399994 1200.0 Late delivery \n", "\n", " Late_delivery_risk Category Name Customer City Customer Country ... \\\n", "172 1 Computers Caguas Puerto Rico ... \n", "1353 0 Computers Caguas Puerto Rico ... \n", "1370 0 Computers Caguas Puerto Rico ... \n", "1375 1 Computers Caguas Puerto Rico ... \n", "1594 1 Computers Caguas Puerto Rico ... \n", "... ... ... ... ... ... \n", "170177 0 Computers Jacksonville EE. UU. ... \n", "170308 0 Computers New York EE. UU. ... \n", "170841 0 Computers Algonquin EE. UU. ... \n", "173423 0 Computers Caguas Puerto Rico ... \n", "173932 1 Computers Caguas Puerto Rico ... \n", "\n", " Order Item Product Price Order Item Profit Ratio Order Item Quantity \\\n", "172 1500.0 0.08 1 \n", "1353 1500.0 0.07 1 \n", "1370 1500.0 0.47 1 \n", "1375 1500.0 0.33 1 \n", "1594 1500.0 0.25 1 \n", "... ... ... ... \n", "170177 1500.0 0.31 1 \n", "170308 1500.0 0.38 1 \n", "170841 1500.0 0.13 1 \n", "173423 1500.0 0.10 1 \n", "173932 1500.0 -0.27 1 \n", "\n", " Sales Order Region Order State Order Status \\\n", "172 1500.0 Northern Europe Inglaterra COMPLETE \n", "1353 1500.0 Western Europe Isla de Francia PAYMENT_REVIEW \n", "1370 1500.0 Southern Europe Madrid PENDING_PAYMENT \n", "1375 1500.0 Western Europe Utrecht PENDING_PAYMENT \n", "1594 1500.0 Western Europe Viena ON_HOLD \n", "... ... ... ... ... \n", "170177 1500.0 Western Europe Berlín PROCESSING \n", "170308 1500.0 Northern Europe Inglaterra SUSPECTED_FRAUD \n", "170841 1500.0 South Asia Daca PROCESSING \n", "173423 1500.0 Northern Europe Dublín CLOSED \n", "173932 1500.0 Eastern Asia Heilongjiang CLOSED \n", "\n", " Product Name shipping date (DateOrders) Shipping Mode \n", "172 Dell Laptop 2017-10-31 18:21:00 Second Class \n", "1353 Dell Laptop 2017-10-30 19:55:00 Standard Class \n", "1370 Dell Laptop 2017-10-31 17:07:00 Standard Class \n", "1375 Dell Laptop 2017-11-03 16:25:00 Standard Class \n", "1594 Dell Laptop 2017-11-05 10:17:00 Second Class \n", "... ... ... ... \n", "170177 Dell Laptop 2017-10-31 03:06:00 Standard Class \n", "170308 Dell Laptop 2017-10-30 12:44:00 Standard Class \n", "170841 Dell Laptop 2017-12-27 06:32:00 Standard Class \n", "173423 Dell Laptop 2017-10-30 19:45:00 Standard Class \n", "173932 Dell Laptop 2017-12-30 05:50:00 Standard Class \n", "\n", "[158 rows x 32 columns]\n" ] } ] }, { "cell_type": "code", "source": [ "print(df.columns.tolist())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "5WIZ3RNhRvn-", "outputId": "9948b8bf-6313-4ff0-bf8f-45ad589caf9f" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "['Type', 'Days for shipping (real)', 'Days for shipment (scheduled)', 'Benefit per order', 'Sales per customer', 'Delivery Status', 'Late_delivery_risk', 'Category Name', 'Customer City', 'Customer Country', 'Customer Segment', 'Customer State', 'Customer Zipcode', 'Department Name', 'Latitude', 'Longitude', 'Market', 'Order City', 'Order Country', 'order date (DateOrders)', 'Order Item Discount', 'Order Item Discount Rate', 'Order Item Product Price', 'Order Item Profit Ratio', 'Order Item Quantity', 'Sales', 'Order Region', 'Order State', 'Order Status', 'Product Name', 'shipping date (DateOrders)', 'Shipping Mode']\n" ] } ] }, { "cell_type": "code", "source": [ "print(f\"Duplicate rows: {df.duplicated().sum()}\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "AE_PqzYQSClY", "outputId": "677a35ab-92fa-41ae-cc03-8c08ed213ef8" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Duplicate rows: 0\n" ] } ] }, { "cell_type": "markdown", "source": [ "### Data Cleaning Summary\n", "\n", "The raw dataset contained 180,519 rows and 53 columns. After cleaning, we have 180,516 rows and 32 columns.\n", "\n", "**What was removed:**\n", "- Personal identifiable information: customer name, email, password, and street address\n", "- Columns with excessive missing values: `Order Zipcode` (86% missing) and `Product Description` (100% missing)\n", "- Redundant ID columns that carry no predictive value\n", "- Duplicate columns: `Order Profit Per Order`, `Order Item Total`, and `Product Price` were identical to existing columns under different names\n", "- `Product Status` was a constant column (all zeros)\n", "- 3 rows with missing `Customer Zipcode` values were dropped\n", "\n", "**What was fixed:**\n", "- `order date (DateOrders)` and `shipping date (DateOrders)` were converted from plain text to proper datetime format\n", "- 158 rows with high discount values (up to 500 USD) were investigated and confirmed as valid. They are all Dell Laptop orders with a product price of 1,500 USD, making the discount large in absolute terms but reasonable in context.\n", "\n", "**Note for modeling:**\n", "`Delivery Status` and `Days for shipping (real)` are kept for EDA but will be dropped before model training, as they are outcomes not available at the time of order placement." ], "metadata": { "id": "fBOdLP9zSLCb" } }, { "cell_type": "markdown", "source": [ "#Research" ], "metadata": { "id": "leaiawIDUNAH" } }, { "cell_type": "markdown", "source": [ "### **Research:** Pose relevant questions about your dataset, then answer them using visual elements (e.g. charts or plots) to provide clear insights.\n", "\n", "For example, in the 2nd lecture the entire class took a survey. Then, we talked about the collected data and desplayed the collected data using the right **plots** - Lines, Bars, Hist, Pie, Map, HeatMap, Area, Time, etc.\n", "\n", "An aditional more specific example, would be the questions we asked during the recitation on the `Titanic` dataset:\n", " - \"Did survival rates differ by gender?\"\n", " - \"Was passenger class related to survival?\"\n", " - \"What was the age distribution of survivors vs. non-survivors?\"\n", " - \"Did embarking location (port) have any effect on survival?\" \n", " \n", "And how we answered those questions using **plots**.\n", "\n", "The idea is to pose questions that can uncover patterns, correlations, or anomalies in your dataset, then back those up with clean, insightful visualizations." ], "metadata": { "id": "lo68PsjTK0_j" } }, { "cell_type": "markdown", "source": [ "### Research Questions\n", "\n", "Before diving into the data, we defined 6 questions that each target a different angle of the supply chain. For each question we also wrote down what common sense would predict — then let the data confirm or challenge it.\n", "\n", "**Q1. Which shipping mode has the highest late delivery rate?**\n", "Assumption: Same Day and First Class should have the lowest late delivery rates since they are the premium options customers pay more for.\n", "\n", "**Q2. How has order volume evolved over time?**\n", "Assumption: A growing business — steady upward trend in orders from 2015 to 2018, with profit growing alongside it.\n", "\n", "**Q3. Which global market has the highest late delivery risk?**\n", "Assumption: Markets that are geographically far from the US warehouse — such as Southeast Asia or Africa — should have the highest late delivery rates due to longer and more complex logistics chains.\n", "\n", "**Q4. Does a higher discount rate lead to lower profit?**\n", "Assumption: The higher the discount, the lower the profit — a straightforward negative relationship.\n", "\n", "**Q5. What does the profit distribution look like — how many orders are actually losses?**\n", "Assumption: Most orders are profitable. Losses should be rare edge cases.\n", "\n", "**Q6. Which shipping mode best sticks to its promised schedule?**\n", "Assumption: Same Day shipping should have the smallest gap between scheduled and actual delivery, since it is the most urgent and closely monitored." ], "metadata": { "id": "ChWDKPCzkMRR" } }, { "cell_type": "markdown", "source": [ "\n", "\n", "To better understand the data before building any models, we posed 6 questions across different dimensions of the supply chain: shipping performance, order profitability, geographic patterns, and time trends. For each question we stated an assumption upfront, then let the data answer it.\n", "\n", "The goal was to uncover patterns that would guide our modeling decisions, particularly around what drives late deliveries and what makes an order profitable or not." ], "metadata": { "id": "W_3DiexHhIJ-" } }, { "cell_type": "code", "source": [ "import plotly.express as px\n", "\n", "late_by_mode = (\n", " df.groupby('Shipping Mode')['Late_delivery_risk']\n", " .mean()\n", " .mul(100)\n", " .round(1)\n", " .reset_index()\n", ")\n", "late_by_mode.columns = ['Shipping Mode', 'Late Delivery Rate (%)']\n", "late_by_mode = late_by_mode.sort_values('Late Delivery Rate (%)', ascending=False)\n", "\n", "fig = px.bar(\n", " late_by_mode,\n", " x='Shipping Mode',\n", " y='Late Delivery Rate (%)',\n", " color='Late Delivery Rate (%)',\n", " color_continuous_scale='RdYlGn_r',\n", " text='Late Delivery Rate (%)',\n", " title='Late Delivery Rate by Shipping Mode'\n", ")\n", "fig.update_traces(texttemplate='%{text}%', textposition='outside')\n", "fig.update_layout(\n", " plot_bgcolor='white',\n", " coloraxis_showscale=False,\n", " yaxis=dict(range=[0, 110]),\n", " title_font_size=18\n", ")\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 542 }, "id": "m7nHpwrckWXu", "outputId": "8715fe1e-74a4-489f-efe5-dae3bf8571d3" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "\n", "**Q1 Finding:** First Class has by far the highest late delivery rate at 95.3%, followed by Second Class at 76.6%, Same Day at 45.7%, and Standard Class at the lowest with 38.1%.\n", "\n", "**vs. Our Assumption:** Completely wrong. We assumed premium shipping modes like First Class and Same Day would perform best — instead First Class is the worst offender by a wide margin. Standard Class, the slowest and cheapest option, is actually the most reliable. This suggests that faster shipping promises are being systematically over-committed and under-delivered, which is exactly the kind of operational problem a predictive model could help flag in advance." ], "metadata": { "id": "WHlW2r87rBHz" } }, { "cell_type": "code", "source": [ "from plotly.subplots import make_subplots\n", "import plotly.graph_objects as go\n", "\n", "df['order_month'] = df['order date (DateOrders)'].dt.to_period('M').astype(str)\n", "\n", "monthly = df.groupby('order_month').agg(\n", " order_count=('Late_delivery_risk', 'count'),\n", " avg_profit=('Benefit per order', 'mean')\n", ").reset_index()\n", "\n", "fig = make_subplots(specs=[[{\"secondary_y\": True}]])\n", "\n", "fig.add_trace(\n", " go.Scatter(\n", " x=monthly['order_month'], y=monthly['order_count'],\n", " name='Order Count', fill='tozeroy',\n", " line=dict(color='#636EFA', width=2)\n", " ),\n", " secondary_y=False\n", ")\n", "fig.add_trace(\n", " go.Scatter(\n", " x=monthly['order_month'], y=monthly['avg_profit'],\n", " name='Avg Profit (USD)', line=dict(color='#EF553B', width=2, dash='dash')\n", " ),\n", " secondary_y=True\n", ")\n", "\n", "fig.update_layout(\n", " title='Monthly Order Volume and Average Profit (2015-2018)',\n", " plot_bgcolor='white',\n", " xaxis=dict(tickangle=45, tickfont=dict(size=10)),\n", " title_font_size=18,\n", " legend=dict(x=0.01, y=0.99)\n", ")\n", "fig.update_yaxes(title_text='Order Count', secondary_y=False)\n", "fig.update_yaxes(title_text='Avg Profit (USD)', secondary_y=True)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 542 }, "id": "62AVaSlFrMGi", "outputId": "9a579d9b-123c-4806-9972-4d4a5aeb3da4" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "**Q2 Finding:** Order volume was remarkably flat throughout the entire period, consistently around 5,000 orders per month from 2015 through late 2017 with no real growth trend. Average profit per order also stayed stable around $20-22 USD. The sharp drop in volume and spike in average profit at the end of the chart are not a real business event: they are a data artifact caused by January 2018 being an incomplete month in the dataset.\n", "\n", "**vs. Our Assumption:** Partially wrong. We expected steady growth, but the business appears to have been operating at a plateau rather than expanding. Volume is consistent and predictable, which is actually useful for modeling; but the lack of growth is a signal worth noting. A business with flat order volume over 3 years may be facing saturation or operational constraints." ], "metadata": { "id": "pNSljCBv12DE" } }, { "cell_type": "code", "source": [ "for col in ['Customer Country', 'Order Country', 'Market', 'Order Region', 'Order State']:\n", " print(f\"\\n--- {col} ---\")\n", " print(df[col].unique())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "collapsed": true, "id": "u6Te2vrJciq9", "outputId": "618a0469-70a7-4f1a-8850-c54d20b03ec8" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\n", "--- Customer Country ---\n", "['Puerto Rico' 'EE. UU.']\n", "\n", "--- Order Country ---\n", "['Indonesia' 'India' 'Australia' 'China' 'Japón' 'Corea del Sur'\n", " 'Singapur' 'Turquía' 'Mongolia' 'Estados Unidos' 'Nigeria'\n", " 'República Democrática del Congo' 'Senegal' 'Marruecos' 'Alemania'\n", " 'Francia' 'Países Bajos' 'Reino Unido' 'Guatemala' 'El Salvador' 'Panamá'\n", " 'República Dominicana' 'Venezuela' 'Colombia' 'Honduras' 'Brasil'\n", " 'México' 'Uruguay' 'Argentina' 'Cuba' 'Perú' 'Nicaragua' 'Ecuador'\n", " 'Angola' 'Sudán' 'Somalia' 'Costa de Marfil' 'Egipto' 'Italia' 'España'\n", " 'Suecia' 'Austria' 'Canada' 'Madagascar' 'Argelia' 'Liberia' 'Zambia'\n", " 'Níger' 'SudAfrica' 'Mozambique' 'Tanzania' 'Ruanda' 'Israel'\n", " 'Nueva Zelanda' 'Bangladés' 'Tailandia' 'Irak' 'Arabia Saudí' 'Filipinas'\n", " 'Kazajistán' 'Irán' 'Myanmar (Birmania)' 'Uzbekistán' 'Benín' 'Camerún'\n", " 'Kenia' 'Togo' 'Ucrania' 'Polonia' 'Portugal' 'Rumania'\n", " 'Trinidad y Tobago' 'Afganistán' 'Pakistán' 'Vietnam' 'Malasia'\n", " 'Finlandia' 'Rusia' 'Irlanda' 'Noruega' 'Eslovaquia' 'Bélgica' 'Bolivia'\n", " 'Chile' 'Jamaica' 'Yemen' 'Ghana' 'Guinea' 'Etiopía' 'Bulgaria'\n", " 'Kirguistán' 'Georgia' 'Nepal' 'Emiratos Árabes Unidos' 'Camboya'\n", " 'Uganda' 'Lesoto' 'Lituania' 'Suiza' 'Hungría' 'Dinamarca' 'Haití'\n", " 'Bielorrusia' 'Croacia' 'Laos' 'Baréin' 'Macedonia' 'República Checa'\n", " 'Sri Lanka' 'Zimbabue' 'Eritrea' 'Burkina Faso' 'Costa Rica' 'Libia'\n", " 'Barbados' 'Tayikistán' 'Siria' 'Guadalupe' 'Papúa Nueva Guinea'\n", " 'Azerbaiyán' 'Turkmenistán' 'Paraguay' 'Jordania' 'Hong Kong' 'Martinica'\n", " 'Moldavia' 'Qatar' 'Mali' 'Albania' 'República del Congo'\n", " 'Bosnia y Herzegovina' 'Omán' 'Túnez' 'Sierra Leona' 'Yibuti' 'Burundi'\n", " 'Montenegro' 'Gabón' 'Sudán del Sur' 'Luxemburgo' 'Namibia' 'Mauritania'\n", " 'Grecia' 'Suazilandia' 'Guyana' 'Guayana Francesa'\n", " 'República Centroafricana' 'Taiwán' 'Estonia' 'Líbano' 'Chipre'\n", " 'Guinea-Bissau' 'Surinam' 'Belice' 'Eslovenia' 'República de Gambia'\n", " 'Botsuana' 'Armenia' 'Guinea Ecuatorial' 'Kuwait' 'Bután' 'Chad' 'Serbia'\n", " 'Sáhara Occidental']\n", "\n", "--- Market ---\n", "['Pacific Asia' 'USCA' 'Africa' 'Europe' 'LATAM']\n", "\n", "--- Order Region ---\n", "['Southeast Asia' 'South Asia' 'Oceania' 'Eastern Asia' 'West Asia'\n", " 'West of USA ' 'US Center ' 'West Africa' 'Central Africa' 'North Africa'\n", " 'Western Europe' 'Northern Europe' 'Central America' 'Caribbean'\n", " 'South America' 'East Africa' 'Southern Europe' 'East of USA' 'Canada'\n", " 'Southern Africa' 'Central Asia' 'Eastern Europe' 'South of USA ']\n", "\n", "--- Order State ---\n", "['Java Occidental' 'Rajastán' 'Queensland' ... 'Bistrita-Nasaud' 'Tottori'\n", " 'Khorezm']\n" ] } ] }, { "cell_type": "code", "source": [ "country_translations = {\n", " 'EE. UU.': 'United States', 'Estados Unidos': 'United States',\n", " 'Japón': 'Japan', 'Corea del Sur': 'South Korea', 'Singapur': 'Singapore',\n", " 'Turquía': 'Turkey', 'República Democrática del Congo': 'Democratic Republic of the Congo',\n", " 'Marruecos': 'Morocco', 'Alemania': 'Germany', 'Francia': 'France',\n", " 'Países Bajos': 'Netherlands', 'Reino Unido': 'United Kingdom',\n", " 'República Dominicana': 'Dominican Republic', 'Brasil': 'Brazil',\n", " 'México': 'Mexico', 'Perú': 'Peru', 'Costa de Marfil': 'Ivory Coast',\n", " 'Italia': 'Italy', 'España': 'Spain', 'Suecia': 'Sweden',\n", " 'Argelia': 'Algeria', 'Níger': 'Niger', 'SudAfrica': 'South Africa',\n", " 'Bangladés': 'Bangladesh', 'Tailandia': 'Thailand',\n", " 'Arabia Saudí': 'Saudi Arabia', 'Filipinas': 'Philippines',\n", " 'Kazajistán': 'Kazakhstan', 'Irán': 'Iran', 'Myanmar (Birmania)': 'Myanmar',\n", " 'Uzbekistán': 'Uzbekistan', 'Benín': 'Benin', 'Ucrania': 'Ukraine',\n", " 'Polonia': 'Poland', 'Rumania': 'Romania',\n", " 'Trinidad y Tobago': 'Trinidad and Tobago', 'Afganistán': 'Afghanistan',\n", " 'Pakistán': 'Pakistan', 'Malasia': 'Malaysia', 'Rusia': 'Russia',\n", " 'Irlanda': 'Ireland', 'Noruega': 'Norway', 'Eslovaquia': 'Slovakia',\n", " 'Bélgica': 'Belgium', 'Etiopía': 'Ethiopia', 'Kirguistán': 'Kyrgyzstan',\n", " 'Emiratos Árabes Unidos': 'United Arab Emirates', 'Camboya': 'Cambodia',\n", " 'Lituania': 'Lithuania', 'Suiza': 'Switzerland', 'Hungría': 'Hungary',\n", " 'Dinamarca': 'Denmark', 'Haití': 'Haiti', 'Bielorrusia': 'Belarus',\n", " 'Croacia': 'Croatia', 'Baréin': 'Bahrain', 'Macedonia': 'North Macedonia',\n", " 'República Checa': 'Czech Republic', 'Zimbabue': 'Zimbabwe',\n", " 'Tayikistán': 'Tajikistan', 'Guadalupe': 'Guadeloupe',\n", " 'Papúa Nueva Guinea': 'Papua New Guinea', 'Azerbaiyán': 'Azerbaijan',\n", " 'Turkmenistán': 'Turkmenistan', 'Jordania': 'Jordan',\n", " 'Martinica': 'Martinique', 'Moldavia': 'Moldova',\n", " 'República del Congo': 'Republic of the Congo',\n", " 'Bosnia y Herzegovina': 'Bosnia and Herzegovina',\n", " 'Túnez': 'Tunisia', 'Sierra Leona': 'Sierra Leone',\n", " 'Sudán del Sur': 'South Sudan', 'Luxemburgo': 'Luxembourg',\n", " 'Grecia': 'Greece', 'Suazilandia': 'Eswatini',\n", " 'Guayana Francesa': 'French Guiana',\n", " 'República Centroafricana': 'Central African Republic',\n", " 'Taiwán': 'Taiwan', 'Líbano': 'Lebanon', 'Chipre': 'Cyprus',\n", " 'Surinam': 'Suriname', 'Belice': 'Belize', 'Eslovenia': 'Slovenia',\n", " 'República de Gambia': 'Gambia', 'Botsuana': 'Botswana',\n", " 'Guinea Ecuatorial': 'Equatorial Guinea', 'Bután': 'Bhutan',\n", " 'Sáhara Occidental': 'Western Sahara', 'Gabón': 'Gabon',\n", " 'Yibuti': 'Djibouti', 'Ruanda': 'Rwanda', 'Kenia': 'Kenya',\n", " 'Camerún': 'Cameroon', 'Omán': 'Oman', 'Nueva Zelanda': 'New Zealand',\n", " 'Sudán': 'Sudan', 'Namibia': 'Namibia', 'Mauritania': 'Mauritania',\n", " 'Lesoto': 'Lesotho'\n", "}\n", "\n", "df['Customer Country'] = df['Customer Country'].replace(country_translations)\n", "df['Order Country'] = df['Order Country'].replace(country_translations)\n", "\n", "print(\"Remaining non-ASCII in Order Country:\")\n", "print([c for c in df['Order Country'].unique() if not c.isascii()])" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "mEFrYozPc-8Q", "outputId": "29a04255-d1df-4efa-e651-88ae8a8ca925" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Remaining non-ASCII in Order Country:\n", "['Panamá']\n" ] } ] }, { "cell_type": "code", "source": [ "import plotly.express as px\n", "\n", "map_data = df.groupby('Order Country')['Late_delivery_risk'].mean().reset_index()\n", "map_data.columns = ['Country', 'Late Delivery Rate']\n", "map_data['Late Delivery Rate (%)'] = (map_data['Late Delivery Rate'] * 100).round(1)\n", "\n", "fig = px.choropleth(\n", " map_data,\n", " locations='Country',\n", " locationmode='country names',\n", " color='Late Delivery Rate (%)',\n", " color_continuous_scale='RdYlGn_r',\n", " title='Late Delivery Risk by Country',\n", " labels={'Late Delivery Rate (%)': 'Late Delivery Rate (%)'},\n", ")\n", "\n", "fig.update_layout(\n", " title_font_size=22,\n", " geo=dict(showframe=False, showcoastlines=True, projection_type='natural earth'),\n", " coloraxis_colorbar=dict(title='Late Delivery
Rate (%)'),\n", " margin=dict(l=0, r=0, t=50, b=0),\n", " height=550\n", ")\n", "\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 567 }, "id": "1DzcvUoXdEXG", "outputId": "7c97e6cd-0ab8-4505-b723-8ec467b9e853" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "**Q3 Finding:** Late delivery risk is surprisingly uniform across most of the world, with the majority of countries sitting in the 50-70% range (yellow). The most striking outliers are several African nations showing 100% late delivery rates (deep red), while a small number of countries in Africa and Asia show near-zero rates (green). Grey countries had too few orders to register.\n", "\n", "**vs. Our Assumption:** We expected a clear geographic pattern with developed regions performing better than emerging markets. Instead the map tells a different story: geography alone is not the main driver of late deliveries. The near-global uniformity suggests the problem is systemic and operational, rooted in shipping mode choices and order management rather than destination country. The African extremes on both ends likely reflect very low order volumes making the percentages volatile rather than a genuine regional pattern." ], "metadata": { "id": "Ur--Yjoode2n" } }, { "cell_type": "code", "source": [ "sample = df.sample(5000, random_state=42)\n", "sample = sample.copy()\n", "sample['Order Result'] = sample['Benefit per order'].apply(\n", " lambda x: 'Profit' if x >= 0 else 'Loss'\n", ")\n", "\n", "fig = px.scatter(\n", " sample,\n", " x='Order Item Discount Rate',\n", " y='Benefit per order',\n", " color='Order Result',\n", " color_discrete_map={'Profit': '#00CC96', 'Loss': '#EF553B'},\n", " opacity=0.4,\n", " title='Does a Higher Discount Lead to Lower Profit?',\n", " labels={\n", " 'Order Item Discount Rate': 'Discount Rate',\n", " 'Benefit per order': 'Profit per Order (USD)',\n", " 'Order Result': ''\n", " }\n", ")\n", "fig.add_hline(\n", " y=0, line_dash='dash', line_color='black',\n", " annotation_text='Break-even', annotation_position='top right'\n", ")\n", "fig.update_layout(plot_bgcolor='white', title_font_size=18)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 542 }, "id": "2c7gU1uI19-t", "outputId": "1a367d31-ce03-4d0c-c793-f41e969f93ba" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "**Q4 Finding:** Discount rate alone does not determine whether an order is profitable. At every single discount level, including 0%, there are both profitable (green) and loss-making (red) orders. Losses appear consistently across the entire discount range, and some of the deepest losses actually occur at mid-range discounts around 0.05 to 0.15. Even at the maximum 25% discount, profitable orders still exist.\n", "\n", "**vs. Our Assumption:** We expected a clear downward trend where higher discounts mean lower profit. The chart shows the opposite: profitability is spread across all discount levels with no meaningful pattern. This tells us that discount rate is not the key driver of profit or loss. Other factors like product type, shipping cost, and order size likely matter far more, which makes this an important signal for our regression model." ], "metadata": { "id": "7F8CHEmadtjf" } }, { "cell_type": "code", "source": [ "import plotly.graph_objects as go\n", "import numpy as np\n", "from scipy.stats import gaussian_kde\n", "\n", "# KDE calculation\n", "data = df['Benefit per order'].dropna()\n", "kde = gaussian_kde(data, bw_method=0.3)\n", "x_range = np.linspace(data.min(), data.max(), 1000)\n", "y_kde = kde(x_range)\n", "\n", "# split into loss and profit\n", "x_loss = x_range[x_range <= 0]\n", "y_loss = y_kde[x_range <= 0]\n", "x_profit = x_range[x_range >= 0]\n", "y_profit = y_kde[x_range >= 0]\n", "\n", "fig = go.Figure()\n", "\n", "# loss area\n", "fig.add_trace(go.Scatter(\n", " x=np.concatenate([x_loss, [0]]),\n", " y=np.concatenate([y_loss, [kde(np.array([0]))[0]]]),\n", " fill='tozeroy', mode='none',\n", " fillcolor='rgba(255, 99, 71, 0.5)',\n", " name='Loss'\n", "))\n", "\n", "# profit area\n", "fig.add_trace(go.Scatter(\n", " x=np.concatenate([[0], x_profit]),\n", " y=np.concatenate([[kde(np.array([0]))[0]], y_profit]),\n", " fill='tozeroy', mode='none',\n", " fillcolor='rgba(100, 200, 150, 0.5)',\n", " name='Profit'\n", "))\n", "\n", "# break-even line\n", "fig.add_vline(x=0, line_dash='dash', line_color='black', line_width=2)\n", "\n", "fig.update_layout(\n", " title='Distribution of Profit per Order — How Many Orders Are Losses?',\n", " xaxis_title='Profit per Order (USD)',\n", " yaxis_title='Density',\n", " height=500,\n", " annotations=[\n", " dict(x=-500, y=max(y_kde)*0.6, text='18.7% of orders
are losses',\n", " showarrow=False, font=dict(color='crimson', size=13)),\n", " dict(x=300, y=max(y_kde)*0.6, text='81.3% of orders
are profitable',\n", " showarrow=False, font=dict(color='seagreen', size=13))\n", " ]\n", ")\n", "\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 517 }, "id": "20Pg_ndkfIFa", "outputId": "553798e3-cd32-4f94-9a8a-6eb50df8c835" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "**Q5 Finding:** 81.3% of orders are profitable while 18.7% result in a loss. The density curve shows the distribution is heavily concentrated just above zero, meaning most profits are small and modest. The loss side tells a different story while fewer in count, losses stretch far to the left reaching to around 4,000 USD, showing that when orders go wrong they can go very wrong.\n", "\n", "**vs. Our Assumption:** We expected losses to be rare edge cases. Nearly 1 in 5 orders losing money is a bigger problem than expected. The asymmetry between small typical profits and large extreme losses makes profit prediction a genuinely valuable modeling target for the business." ], "metadata": { "id": "xO_ecKWnfgQ6" } }, { "cell_type": "code", "source": [ "df['Shipping Gap'] = df['Days for shipping (real)'] - df['Days for shipment (scheduled)']\n", "\n", "fig = px.box(\n", " df,\n", " x='Shipping Mode',\n", " y='Shipping Gap',\n", " color='Shipping Mode',\n", " title='How Well Does Each Shipping Mode Stick to Its Schedule?',\n", " labels={\n", " 'Shipping Gap': 'Days Late (Actual minus Scheduled)',\n", " 'Shipping Mode': 'Shipping Mode'\n", " },\n", " points=False,\n", " color_discrete_sequence=px.colors.qualitative.Safe\n", ")\n", "fig.add_hline(\n", " y=0, line_dash='dash', line_color='black',\n", " annotation_text='On Schedule', annotation_position='top right'\n", ")\n", "fig.update_layout(\n", " plot_bgcolor='white',\n", " showlegend=False,\n", " title_font_size=18\n", ")\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 542 }, "id": "twLfGdAo2TBs", "outputId": "9d7550ab-bd3e-454c-d67a-75b621f41974" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "**Q6 Finding:** Second Class is the worst performing shipping mode, with a median of 2 days late and a range stretching up to 4 days behind schedule. First Class has almost no spread at all, with nearly every order landing exactly 1 day late, which explains the 95.3% late delivery rate we saw in Q1. Standard Class is the most balanced, centered right around the schedule line with roughly equal early and late deliveries. Same Day is the most reliable, staying within 1 day of schedule in both directions.\n", "\n", "**vs. Our Assumption:** We expected Same Day and First Class to be the most reliable since customers pay more for them. Same Day confirmed that expectation but First Class completely contradicted it. A shipping mode that almost universally arrives exactly 1 day late points to a systematic scheduling problem rather than random delays. The scheduled times for First Class orders are likely being set unrealistically tight across the board, which is something a predictive model could help the business catch early." ], "metadata": { "id": "DixN4hXLgmJM" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "zrIvgbNoIPU6" } }, { "cell_type": "markdown", "source": [ "# **Part 3: Define and Train a baseline model**\n", "\n", "1. **Regression Goal**: Clearly state the problem you’re addressing.\n", "\n", "2. **Feature Selection**: Identify the features that seem most relevant. It’s fine to start with all features if you’re unsure.\n", "\n", "3. **Train-Test Split**: Partition your data into training, and testing sets. Use simple sampling. Quick reminder - when using ramdom - Use `Seed`!\n", "\n", "4. **Model Training**: For simplicity, start with default parameters on a `Linear Regression` model, using scikit-learn. Focus on establishing a baseline.\n", "\n", "5. **Model Evaluation**: Present straightforward metrics such as MAE, MSE, RMSE, R2, etc.\n", "\n", "6. **Insights**: Summarize the model’s performance with visuals.\n", "\n", "7. **Feature Importance:** Explain & Visualize `feature importance` by looking on the `coefficients` of the Linear Regression model .\n", "\n", "
\n", "\n", "*FYI: Sections 5 and 6 will be repeated throughout your work.*" ], "metadata": { "id": "TxKHPqppIPZT" } }, { "cell_type": "markdown", "source": [ "### Part 3: Baseline Regression Model\n", "\n", "**Goal:** Predict `Benefit per order`: the profit or loss in USD for each shipment, using only information available at the time an order is placed.\n", "\n", "This baseline uses a simple Linear Regression model with no feature engineering and default parameters. The purpose is not to get a great result, but to set a performance benchmark that our improved models in Part 5 will need to beat." ], "metadata": { "id": "EfniaMYBK6sK" } }, { "cell_type": "code", "source": [ "# Step 1: Select features and prepare the dataset for modeling\n", "# We drop columns that are either outcomes (leaky) or too granular for a baseline\n", "baseline_drop = [\n", " 'Delivery Status',\n", " 'Days for shipping (real)',\n", " 'Shipping Gap',\n", " 'Late_delivery_risk',\n", " 'Order Item Profit Ratio',\n", " 'order date (DateOrders)',\n", " 'shipping date (DateOrders)',\n", " 'Customer City', 'Customer State', 'Customer Country',\n", " 'Customer Zipcode', 'Order City', 'Order State',\n", " 'Order Country', 'Latitude', 'Longitude',\n", " 'Order Status', 'Product Name'\n", "]\n", "\n", "df_model = df.drop(columns=baseline_drop)\n", "\n", "# encode categorical text columns into numbers so sklearn can read them\n", "df_model = pd.get_dummies(df_model, drop_first=True)\n", "\n", "print(df_model.shape)\n", "print(df_model.columns.tolist())" ], "metadata": { "id": "euWXtGKHK65d", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "2bfc1a82-2c37-40ed-a7ad-7f3a3ce3471c" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180516, 137)\n", "['Days for shipment (scheduled)', 'Benefit per order', 'Sales per customer', 'Order Item Discount', 'Order Item Discount Rate', 'Order Item Product Price', 'Order Item Quantity', 'Sales', 'Type_DEBIT', 'Type_PAYMENT', 'Type_TRANSFER', 'Category Name_As Seen on TV!', 'Category Name_Baby ', 'Category Name_Baseball & Softball', 'Category Name_Basketball', 'Category Name_Books ', 'Category Name_Boxing & MMA', 'Category Name_CDs ', 'Category Name_Cameras ', 'Category Name_Camping & Hiking', 'Category Name_Cardio Equipment', \"Category Name_Children's Clothing\", 'Category Name_Cleats', 'Category Name_Computers', 'Category Name_Consumer Electronics', 'Category Name_Crafts', 'Category Name_DVDs', 'Category Name_Electronics', 'Category Name_Fishing', 'Category Name_Fitness Accessories', 'Category Name_Garden', \"Category Name_Girls' Apparel\", 'Category Name_Golf Apparel', 'Category Name_Golf Bags & Carts', 'Category Name_Golf Balls', 'Category Name_Golf Gloves', 'Category Name_Golf Shoes', 'Category Name_Health and Beauty', 'Category Name_Hockey', 'Category Name_Hunting & Shooting', 'Category Name_Indoor/Outdoor Games', \"Category Name_Kids' Golf Clubs\", 'Category Name_Lacrosse', \"Category Name_Men's Clothing\", \"Category Name_Men's Footwear\", \"Category Name_Men's Golf Clubs\", 'Category Name_Music', 'Category Name_Pet Supplies', 'Category Name_Shop By Sport', 'Category Name_Soccer', 'Category Name_Sporting Goods', 'Category Name_Strength Training', 'Category Name_Tennis & Racquet', 'Category Name_Toys', 'Category Name_Trade-In', 'Category Name_Video Games', 'Category Name_Water Sports', \"Category Name_Women's Apparel\", \"Category Name_Women's Clothing\", \"Category Name_Women's Golf Clubs\", 'Customer Segment_Corporate', 'Customer Segment_Home Office', 'Department Name_Book Shop', 'Department Name_Discs Shop', 'Department Name_Fan Shop', 'Department Name_Fitness', 'Department Name_Footwear', 'Department Name_Golf', 'Department Name_Health and Beauty ', 'Department Name_Outdoors', 'Department Name_Pet Shop', 'Department Name_Technology', 'Market_Europe', 'Market_LATAM', 'Market_Pacific Asia', 'Market_USCA', 'Order Region_Caribbean', 'Order Region_Central Africa', 'Order Region_Central America', 'Order Region_Central Asia', 'Order Region_East Africa', 'Order Region_East of USA', 'Order Region_Eastern Asia', 'Order Region_Eastern Europe', 'Order Region_North Africa', 'Order Region_Northern Europe', 'Order Region_Oceania', 'Order Region_South America', 'Order Region_South Asia', 'Order Region_South of USA ', 'Order Region_Southeast Asia', 'Order Region_Southern Africa', 'Order Region_Southern Europe', 'Order Region_US Center ', 'Order Region_West Africa', 'Order Region_West Asia', 'Order Region_West of USA ', 'Order Region_Western Europe', 'Shipping Mode_Same Day', 'Shipping Mode_Second Class', 'Shipping Mode_Standard Class', 'order_month_2015-02', 'order_month_2015-03', 'order_month_2015-04', 'order_month_2015-05', 'order_month_2015-06', 'order_month_2015-07', 'order_month_2015-08', 'order_month_2015-09', 'order_month_2015-10', 'order_month_2015-11', 'order_month_2015-12', 'order_month_2016-01', 'order_month_2016-02', 'order_month_2016-03', 'order_month_2016-04', 'order_month_2016-05', 'order_month_2016-06', 'order_month_2016-07', 'order_month_2016-08', 'order_month_2016-09', 'order_month_2016-10', 'order_month_2016-11', 'order_month_2016-12', 'order_month_2017-01', 'order_month_2017-02', 'order_month_2017-03', 'order_month_2017-04', 'order_month_2017-05', 'order_month_2017-06', 'order_month_2017-07', 'order_month_2017-08', 'order_month_2017-09', 'order_month_2017-10', 'order_month_2017-11', 'order_month_2017-12', 'order_month_2018-01']\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 2 - Split the data into training and testing sets\n", "# 80% of the data is used to train the model, 20% is held back to test it\n", "# random_state=SEED ensures the split is the same every time we run this\n", "from sklearn.model_selection import train_test_split\n", "\n", "X = df_model.drop(columns=['Benefit per order'])\n", "y = df_model['Benefit per order']\n", "\n", "X_train, X_test, y_train, y_test = train_test_split(\n", " X, y, test_size=0.2, random_state=SEED\n", ")\n", "\n", "print(f\"Train size: {X_train.shape[0]} rows\")\n", "print(f\"Test size: {X_test.shape[0]} rows\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "uYTgLJNdidE4", "outputId": "bdb2ed17-e6ca-4f41-9ede-afe8aef6bb9a" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train size: 144412 rows\n", "Test size: 36104 rows\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 3 - Train the baseline Linear Regression model\n", "# We use all default parameters: no tuning, no engineering\n", "# This gives us a benchmark number to beat in Part 5\n", "from sklearn.linear_model import LinearRegression\n", "\n", "baseline_model = LinearRegression()\n", "baseline_model.fit(X_train, y_train)\n", "\n", "print(\"Baseline model trained.\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "lxSAzJw3ifJo", "outputId": "ccc4b2c2-ccc8-4eec-e80c-de284a6a70fe" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Baseline model trained.\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 4: Evaluate the model on the test set\n", "# MAE: average error in USD\n", "# RMSE: same unit as target, penalises large errors more\n", "# R²: how much of the variance in profit the model explains (1.0 = perfect)\n", "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n", "import numpy as np\n", "\n", "y_pred = baseline_model.predict(X_test)\n", "\n", "mae = mean_absolute_error(y_test, y_pred)\n", "mse = mean_squared_error(y_test, y_pred)\n", "rmse = np.sqrt(mse)\n", "r2 = r2_score(y_test, y_pred)\n", "\n", "print(f\"MAE: {mae:.2f}\")\n", "print(f\"MSE: {mse:.2f}\")\n", "print(f\"RMSE: {rmse:.2f}\")\n", "print(f\"R²: {r2:.4f}\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "auCjI4LGifGz", "outputId": "0a9df171-89b2-46b3-8383-29003020a173" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "MAE: 53.77\n", "MSE: 10670.03\n", "RMSE: 103.30\n", "R²: 0.0158\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 5: Plot predicted vs actual profit\n", "# Each dot is one order. if the model were perfect all dots would sit on the red line\n", "# The more scattered the dots, the worse the model\n", "import plotly.express as px\n", "import pandas as pd\n", "\n", "results_df = pd.DataFrame({'Actual': y_test, 'Predicted': y_pred})\n", "\n", "fig = px.scatter(\n", " results_df.sample(2000, random_state=SEED),\n", " x='Actual',\n", " y='Predicted',\n", " title='Baseline Model: Predicted vs Actual Profit per Order',\n", " labels={'Actual': 'Actual Profit (USD)', 'Predicted': 'Predicted Profit (USD)'},\n", " opacity=0.4,\n", " color_discrete_sequence=['steelblue']\n", ")\n", "\n", "fig.add_shape(\n", " type='line',\n", " x0=results_df['Actual'].min(), y0=results_df['Actual'].min(),\n", " x1=results_df['Actual'].max(), y1=results_df['Actual'].max(),\n", " line=dict(color='red', dash='dash')\n", ")\n", "\n", "fig.update_layout(height=500)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 517 }, "id": "s-0hppNvijvJ", "outputId": "52ed58f7-2ba9-4124-99eb-b909eb971a62" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "code", "source": [ "# Step 6: Visualise feature importance using model coefficients\n", "# A large positive coefficient means the feature pushes profit up\n", "# A large negative coefficient means it pushes profit down\n", "# We show only the top 15 most influential features\n", "coef_df = pd.DataFrame({\n", " 'Feature': X.columns,\n", " 'Coefficient': baseline_model.coef_\n", "})\n", "\n", "coef_df['Abs'] = coef_df['Coefficient'].abs()\n", "coef_df = coef_df.sort_values('Abs', ascending=False).head(15)\n", "\n", "fig = px.bar(\n", " coef_df,\n", " x='Coefficient',\n", " y='Feature',\n", " orientation='h',\n", " title='Top 15 Feature Coefficients — Baseline Linear Regression',\n", " color='Coefficient',\n", " color_continuous_scale='RdBu',\n", " labels={'Coefficient': 'Coefficient Value', 'Feature': ''}\n", ")\n", "\n", "fig.update_layout(height=500, yaxis=dict(autorange='reversed'))\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 517 }, "id": "mrrqR7YsijtI", "outputId": "919ff171-e070-4c14-abb0-708073202f5c" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "### Baseline Model Results\n", "\n", "| Metric | Value |\n", "|---|---|\n", "| MAE | 53.77 USD |\n", "| MSE | 10,670 |\n", "| RMSE | 103.30 USD |\n", "| R² | 0.016 |\n", "\n", "The baseline Linear Regression explains only 1.6% of the variance in profit per order. On average it is wrong by 53 USD per order, which is significant given that the median profit is only around 31 USD.\n", "\n", "This result is expected. We used raw features with no engineering, no scaling, and a model that assumes linear relationships in data that clearly does not behave linearly (as we saw in the Q4 scatter plot). The R² of 0.016 gives us a clear floor to beat in Part 5. any meaningful feature engineering should produce a dramatic improvement over this baseline." ], "metadata": { "id": "bawxQUXAjaCs" } }, { "cell_type": "markdown", "source": [ "### Part 3 Summary\n", "\n", "The baseline Linear Regression model scores an R² of 0.016, meaning it explains less than 2% of what drives profit per order. On average it is wrong by 53 USD per order, which is poor given the median profit is only around 31 USD.\n", "\n", "This result is expected. Linear Regression assumes straight-line relationships between features and the target, but as we saw in the EDA, profit does not behave that way. The model gives us a clear benchmark to beat.\n", "\n", "In Part 4 we will engineer better features from the existing data, and in Part 5 we will train more powerful models capable of capturing the non-linear patterns the baseline completely missed." ], "metadata": { "id": "JTEAUguqkQra" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "hkerNEsMIPbK" } }, { "cell_type": "markdown", "source": [ "# **Part 4: Feature Engineering**\n", "\n", "* Create, transform, scale, or extract new features; encoding categoricals, polynomial features, PCA, etc.\n", "*TIP: use sklearn's tools, such as Scalar, One-Hot, etc.*\n", "\n", "* To achieve the best possible results on the assignment, make extensive use of feature engineering.\n", "\n", "* Use a `Clustring Model` to create a new feature." ], "metadata": { "id": "wnzIOsrRVc_i" } }, { "cell_type": "code", "source": [ "# Step 1 : Extract date-based features from the order date column\n", "# Day of week, month, quarter and hour can all carry useful signal\n", "# e.g. orders placed late in the month or on weekends may ship differently\n", "\n", "df['order_month'] = df['order date (DateOrders)'].dt.month\n", "df['order_year'] = df['order date (DateOrders)'].dt.year\n", "df['order_dayofweek'] = df['order date (DateOrders)'].dt.dayofweek # 0=Monday, 6=Sunday\n", "df['order_quarter'] = df['order date (DateOrders)'].dt.quarter\n", "df['order_hour'] = df['order date (DateOrders)'].dt.hour\n", "\n", "print(df[['order_month','order_year','order_dayofweek','order_quarter','order_hour']].head())" ], "metadata": { "id": "uQHNkN1nVdTm", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "bfb34762-0c62-4279-8392-3e08b34b1e10" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " order_month order_year order_dayofweek order_quarter order_hour\n", "0 1 2018 2 1 22\n", "1 1 2018 5 1 12\n", "2 1 2018 5 1 12\n", "3 1 2018 5 1 11\n", "4 1 2018 5 1 11\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 2 : Engineer new financial and shipping features\n", "# These capture relationships the raw columns cannot express on their own\n", "\n", "# revenue per item in the order\n", "df['revenue_per_item'] = df['Sales per customer'] / df['Order Item Quantity']\n", "\n", "# how much of the sale price is being discounted\n", "df['discount_to_price_ratio'] = df['Order Item Discount'] / df['Order Item Product Price']\n", "\n", "# flag orders where the discount is aggressive (above 15%)\n", "df['is_high_discount'] = (df['Order Item Discount Rate'] > 0.15).astype(int)\n", "\n", "# flag orders that need to ship fast (2 days or less scheduled)\n", "df['is_urgent'] = (df['Days for shipment (scheduled)'] <= 2).astype(int)\n", "\n", "# flag weekend orders\n", "df['is_weekend'] = (df['order_dayofweek'] >= 5).astype(int)\n", "\n", "print(df[['revenue_per_item','discount_to_price_ratio','is_high_discount','is_urgent','is_weekend']].head())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "1SMQxgpBlpEU", "outputId": "c3676c27-9562-4529-e7eb-1669383bd98f" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " revenue_per_item discount_to_price_ratio is_high_discount is_urgent \\\n", "0 314.640015 0.040000 0 0 \n", "1 311.359985 0.050008 0 0 \n", "2 309.720001 0.055011 0 0 \n", "3 304.809998 0.069992 0 0 \n", "4 298.250000 0.090008 0 0 \n", "\n", " is_weekend \n", "0 0 \n", "1 1 \n", "2 1 \n", "3 1 \n", "4 1 \n" ] } ] }, { "cell_type": "code", "source": [ "# Step 3 : Build the engineered feature matrix\n", "# Drop leaky columns and columns we already extracted info from\n", "# Then encode categoricals and scale numerics\n", "\n", "from sklearn.preprocessing import StandardScaler\n", "\n", "eng_drop = [\n", " 'Delivery Status', 'Days for shipping (real)', 'Shipping Gap',\n", " 'Late_delivery_risk', 'Order Item Profit Ratio',\n", " 'order date (DateOrders)', 'shipping date (DateOrders)',\n", " 'Customer City', 'Customer State', 'Customer Country',\n", " 'Customer Zipcode', 'Order City', 'Order State',\n", " 'Order Country', 'Latitude', 'Longitude',\n", " 'Order Status', 'Product Name'\n", "]\n", "\n", "df_eng = df.drop(columns=eng_drop)\n", "\n", "# encode all categorical columns into numeric 0/1 columns\n", "df_eng = pd.get_dummies(df_eng, drop_first=True)\n", "\n", "# scale all numeric features so no single column dominates due to its unit size\n", "scaler = StandardScaler()\n", "df_eng_scaled = pd.DataFrame(\n", " scaler.fit_transform(df_eng),\n", " columns=df_eng.columns\n", ")\n", "\n", "print(df_eng_scaled.shape)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "qsFV32nhlo5j", "outputId": "26e5b05f-8e51-4dae-bc88-6e125e9a6fea" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "(180516, 111)\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 4 : Find the optimal number of clusters using the elbow method\n", "# We try k from 2 to 10 and look for the point where adding more clusters\n", "# stops meaningfully reducing the inertia (within-cluster variance)\n", "\n", "from sklearn.cluster import KMeans\n", "\n", "# use a sample to speed up the elbow search\n", "sample = df_eng_scaled.sample(10000, random_state=SEED)\n", "\n", "inertias = []\n", "k_range = range(2, 11)\n", "\n", "for k in k_range:\n", " km = KMeans(n_clusters=k, random_state=SEED, n_init=10)\n", " km.fit(sample)\n", " inertias.append(km.inertia_)\n", "\n", "fig = px.line(\n", " x=list(k_range),\n", " y=inertias,\n", " markers=True,\n", " title='Elbow Method : Finding the Right Number of Clusters',\n", " labels={'x': 'Number of Clusters (k)', 'y': 'Inertia'}\n", ")\n", "fig.update_layout(height=400)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 417 }, "id": "PfHrpSnAlo3B", "outputId": "62ae9da3-4616-4b91-9ece-a84d83274f2b" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "code", "source": [ "# Step 5 : Train KMeans with the chosen k\n", "# Pick the k value where the elbow curve bends most sharply\n", "\n", "K = 4 # update this after looking at the elbow plot\n", "\n", "kmeans = KMeans(n_clusters=K, random_state=SEED, n_init=10)\n", "df['cluster'] = kmeans.fit_predict(df_eng_scaled)\n", "\n", "print(df['cluster'].value_counts().sort_index())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "l-hKI-x8lo0e", "outputId": "0c2f4cb0-cc39-48c7-ba89-79e18a766ac3" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "cluster\n", "0 14510\n", "1 51980\n", "2 113569\n", "3 457\n", "Name: count, dtype: int64\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 6 : Visualise clusters using PCA\n", "# PCA reduces all our features down to 2 dimensions so we can plot them\n", "# Each dot is one order, coloured by which cluster it belongs to\n", "\n", "from sklearn.decomposition import PCA\n", "\n", "pca = PCA(n_components=2, random_state=SEED)\n", "components = pca.fit_transform(df_eng_scaled.sample(10000, random_state=SEED))\n", "\n", "pca_df = pd.DataFrame(components, columns=['PC1', 'PC2'])\n", "pca_df['Cluster'] = df['cluster'].sample(10000, random_state=SEED).values.astype(str)\n", "\n", "fig = px.scatter(\n", " pca_df,\n", " x='PC1', y='PC2',\n", " color='Cluster',\n", " title='Customer Order Clusters (PCA Projection)',\n", " labels={'PC1': 'Principal Component 1', 'PC2': 'Principal Component 2'},\n", " opacity=0.5,\n", " color_discrete_sequence=px.colors.qualitative.Safe\n", ")\n", "fig.update_layout(height=520)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 537 }, "id": "yxmCrxIdlxxJ", "outputId": "e93f8ff1-68ff-4a06-f3bb-15df7162d51b" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "code", "source": [ "# Step 7 : Interpret each cluster by looking at its average feature values\n", "# This tells us what makes each cluster distinct from the others\n", "\n", "cluster_profile = df.groupby('cluster')[[\n", " 'Benefit per order', 'Sales per customer',\n", " 'Order Item Discount Rate', 'Days for shipment (scheduled)',\n", " 'Order Item Quantity', 'Order Item Product Price'\n", "]].mean().round(2)\n", "\n", "print(cluster_profile)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "cHURdtkBl0r1", "outputId": "a012fc5f-ff13-4053-f4d0-018f7f662586" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " Benefit per order Sales per customer Order Item Discount Rate \\\n", "cluster \n", "0 28.34 246.25 0.1 \n", "1 32.89 275.62 0.1 \n", "2 15.65 127.96 0.1 \n", "3 150.31 1360.62 0.1 \n", "\n", " Days for shipment (scheduled) Order Item Quantity \\\n", "cluster \n", "0 2.93 2.99 \n", "1 2.93 1.00 \n", "2 2.93 2.54 \n", "3 3.17 1.00 \n", "\n", " Order Item Product Price \n", "cluster \n", "0 92.46 \n", "1 306.66 \n", "2 66.21 \n", "3 1516.41 \n" ] } ] }, { "cell_type": "code", "source": [ "# Step 8 : Add cluster distance to centroid as an extra feature\n", "# How far an order sits from its cluster centre adds useful signal\n", "# Orders far from their centroid are unusual or borderline cases\n", "\n", "from sklearn.metrics import pairwise_distances_argmin_min\n", "\n", "_, distances = pairwise_distances_argmin_min(df_eng_scaled, kmeans.cluster_centers_)\n", "df['cluster_distance'] = distances\n", "\n", "print(df[['cluster', 'cluster_distance']].head())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "p6GEwNHxl1aV", "outputId": "21da9b75-9db0-4a25-aee3-b0e7a490a57b" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " cluster cluster_distance\n", "0 1 24.960023\n", "1 1 25.213549\n", "2 1 25.162019\n", "3 1 25.007403\n", "4 1 25.007339\n" ] } ] }, { "cell_type": "markdown", "source": [ "### Clustering Results : 4 Order Segments\n", "\n", "We used KMeans with k=4, chosen from the elbow plot. The PCA projection confirms the clusters are meaningful.\n", "\n", "| Cluster | Avg Profit | Avg Product Price | Avg Qty | Label |\n", "|---|---|---|---|---|\n", "| 0 | 28 USD | 92 USD | 3.0 | Budget multi-item |\n", "| 1 | 33 USD | 307 USD | 1.0 | Single mid-range item |\n", "| 2 | 16 USD | 66 USD | 2.5 | Low value multi-item |\n", "| 3 | 150 USD | 1,516 USD | 1.0 | High value single item |\n", "\n", "Cluster 3 is the most distinct group, completely separated in the PCA plot. These are high-ticket single item orders (primarily Dell Laptops) that generate 5x more profit than average. The other three clusters are separated mainly by quantity and product price tier.\n", "\n", "Notably, discount rate is identical across all four clusters at 0.10, which reinforces our earlier finding that discounts alone do not drive profitability. The cluster labels and distances to centroid have been added as new features for use in Part 5." ], "metadata": { "id": "qRWosDUBnIxE" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "KLnSsY0UVah_" } }, { "cell_type": "markdown", "source": [ "### Part 5: Three Improved Regression Models\n", "\n", "We now retrain on the full engineered dataset, which includes date features, financial ratios, shipping flags, and the cluster labels from Part 4. We compare three models against the baseline to measure how much the feature engineering actually helped." ], "metadata": { "id": "b1y2QWE5XRIQ" } }, { "cell_type": "code", "source": [ "# Step 1 : Build the final engineered feature matrix\n", "# Same leaky columns dropped as before, but now includes all new features\n", "\n", "eng_drop = [\n", " 'Delivery Status', 'Days for shipping (real)', 'Shipping Gap',\n", " 'Late_delivery_risk', 'Order Item Profit Ratio',\n", " 'order date (DateOrders)', 'shipping date (DateOrders)',\n", " 'Customer City', 'Customer State', 'Customer Country',\n", " 'Customer Zipcode', 'Order City', 'Order State',\n", " 'Order Country', 'Latitude', 'Longitude',\n", " 'Order Status', 'Product Name'\n", "]\n", "\n", "df_final = df.drop(columns=eng_drop)\n", "df_final = pd.get_dummies(df_final, drop_first=True)\n", "\n", "X_eng = df_final.drop(columns=['Benefit per order'])\n", "y_eng = df_final['Benefit per order']\n", "\n", "# scale all features so Linear Regression treats them fairly\n", "scaler2 = StandardScaler()\n", "X_eng_scaled = pd.DataFrame(scaler2.fit_transform(X_eng), columns=X_eng.columns)\n", "\n", "# same 80/20 split as baseline for a fair comparison\n", "X_train_eng, X_test_eng, y_train_eng, y_test_eng = train_test_split(\n", " X_eng_scaled, y_eng, test_size=0.2, random_state=SEED\n", ")\n", "\n", "print(f\"Features : {X_eng_scaled.shape[1]}\")\n", "print(f\"Train rows : {X_train_eng.shape[0]}\")\n", "print(f\"Test rows : {X_test_eng.shape[0]}\")" ], "metadata": { "id": "59VDTUuWXETD", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "fdfd6028-4841-4123-8a5b-490435d1e933" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Features : 112\n", "Train rows : 144412\n", "Test rows : 36104\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 2 : Retrain Linear Regression on the engineered features\n", "# Same model type as the baseline but now with richer inputs\n", "\n", "lr_eng = LinearRegression()\n", "lr_eng.fit(X_train_eng, y_train_eng)\n", "y_pred_lr = lr_eng.predict(X_test_eng)\n", "\n", "mae_lr = mean_absolute_error(y_test_eng, y_pred_lr)\n", "rmse_lr = np.sqrt(mean_squared_error(y_test_eng, y_pred_lr))\n", "r2_lr = r2_score(y_test_eng, y_pred_lr)\n", "\n", "print(f\"Linear Regression (engineered) | MAE: {mae_lr:.2f} RMSE: {rmse_lr:.2f} R²: {r2_lr:.4f}\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "8OfghcZuXWxl", "outputId": "ffabf64c-dd9f-477d-8650-f0b5f247b580" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Linear Regression (engineered) | MAE: 57.36 RMSE: 86.00 R²: 0.3179\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 3 : Train a Random Forest Regressor\n", "# Builds hundreds of decision trees and averages their predictions\n", "# Handles non-linear relationships and interactions between features naturally\n", "\n", "from sklearn.ensemble import RandomForestRegressor\n", "\n", "rf = RandomForestRegressor(n_estimators=100, random_state=SEED, n_jobs=-1)\n", "rf.fit(X_train_eng, y_train_eng)\n", "y_pred_rf = rf.predict(X_test_eng)\n", "\n", "mae_rf = mean_absolute_error(y_test_eng, y_pred_rf)\n", "rmse_rf = np.sqrt(mean_squared_error(y_test_eng, y_pred_rf))\n", "r2_rf = r2_score(y_test_eng, y_pred_rf)\n", "\n", "print(f\"Random Forest | MAE: {mae_rf:.2f} RMSE: {rmse_rf:.2f} R²: {r2_rf:.4f}\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "e0SUbp1HXW_N", "outputId": "a8892ea4-75db-4924-f475-dcad7731bf78" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Random Forest | MAE: 47.79 RMSE: 75.05 R²: 0.4805\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 4 : Train a Gradient Boosting Regressor\n", "# Builds trees sequentially where each tree corrects the errors of the previous one\n", "# Typically the strongest performer on structured tabular data\n", "\n", "from sklearn.ensemble import GradientBoostingRegressor\n", "\n", "gb = GradientBoostingRegressor(n_estimators=100, random_state=SEED)\n", "gb.fit(X_train_eng, y_train_eng)\n", "y_pred_gb = gb.predict(X_test_eng)\n", "\n", "mae_gb = mean_absolute_error(y_test_eng, y_pred_gb)\n", "rmse_gb = np.sqrt(mean_squared_error(y_test_eng, y_pred_gb))\n", "r2_gb = r2_score(y_test_eng, y_pred_gb)\n", "\n", "print(f\"Gradient Boosting | MAE: {mae_gb:.2f} RMSE: {rmse_gb:.2f} R²: {r2_gb:.4f}\")" ], "metadata": { "id": "eLomwhyRXbUY", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "5f025764-3fb1-4279-a56e-2fd4e42c3137" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Gradient Boosting | MAE: 52.86 RMSE: 93.25 R²: 0.1980\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 5 : Compare all models including the baseline in one table and chart\n", "\n", "results = pd.DataFrame({\n", " 'Model': ['Baseline LR', 'Linear Regression', 'Random Forest', 'Gradient Boosting'],\n", " 'MAE': [53.77, mae_lr, mae_rf, mae_gb],\n", " 'RMSE': [103.30, rmse_lr, rmse_rf, rmse_gb],\n", " 'R²': [0.0158, r2_lr, r2_rf, r2_gb]\n", "})\n", "\n", "print(results.to_string(index=False))\n", "\n", "fig = px.bar(\n", " results,\n", " x='Model', y='R²',\n", " title='Model Comparison : R² Score',\n", " color='Model',\n", " text='R²',\n", " color_discrete_sequence=px.colors.qualitative.Safe\n", ")\n", "fig.update_traces(texttemplate='%{text:.3f}', textposition='outside')\n", "fig.update_layout(height=450, showlegend=False, yaxis_range=[0, 1])\n", "fig.show()" ], "metadata": { "id": "n-_qWp15XbcK", "colab": { "base_uri": "https://localhost:8080/", "height": 552 }, "outputId": "cddd64d3-2591-4bfe-a92a-450459e7d8f5" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " Model MAE RMSE R²\n", " Baseline LR 53.770000 103.300000 0.015800\n", "Linear Regression 57.361606 85.995797 0.317878\n", " Random Forest 47.789399 75.047813 0.480502\n", "Gradient Boosting 52.863802 93.247383 0.197988\n" ] }, { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "### Part 5 Results\n", "\n", "| Model | MAE | RMSE | R² |\n", "|---|---|---|---|\n", "| Baseline LR | 53.77 | 103.30 | 0.016 |\n", "| Linear Regression | 57.36 | 86.00 | 0.318 |\n", "| Random Forest | 47.79 | 75.05 | 0.481 |\n", "| Gradient Boosting | 52.86 | 93.25 | 0.198 |\n", "\n", "**Winner : Random Forest**\n", "\n", "Three things stand out from these results.\n", "\n", "First, feature engineering alone had a massive effect. The Linear Regression model went from R² 0.016 to 0.318 using the exact same algorithm, the only difference was better inputs. This confirms that what you feed a model matters far more than which model you pick.\n", "\n", "Second, Random Forest is the clear winner across all three metrics. It explains 48% of the variance in profit per order, reduces average error to 47.79 USD, and has the lowest RMSE at 75.05. Its strength comes from building 100 decision trees and averaging their predictions, which naturally captures the non-linear relationships and feature interactions that Linear Regression cannot.\n", "\n", "Third, Gradient Boosting underperformed Random Forest with default parameters. This is not unusual, Gradient Boosting is more sensitive to hyperparameter settings and typically needs tuning to reach its potential. With Optuna in Part 8 we will push it further." ], "metadata": { "id": "581lKSgFdi8k" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "ykjbsmtgvbLx" } }, { "cell_type": "markdown", "source": [ "# Part 6: Winning Model" ], "metadata": { "id": "PS9t6m0_nJ9t" } }, { "cell_type": "markdown", "source": [ "1. Open a new HuggingFace Model Repository.\n", "2. Export the winning model to a `pickle` file.\n", "3. Upload the pickle file to your new model repository on `HF`." ], "metadata": { "id": "ga-aPfHDnRM3" } }, { "cell_type": "markdown", "source": [ "![image.png](data:image/png;base64,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)" ], "metadata": { "id": "GzsEe2Yun5nC" } }, { "cell_type": "code", "source": [ "# RUN ONCE ONLY : exports and downloads the regression model\n", "# skip this cell after the first run\n", "import pickle\n", "\n", "with open('regression_model.pkl', 'wb') as f:\n", " pickle.dump(rf, f)\n", "\n", "\n", "from google.colab import files\n", "files.download('regression_model.pkl')" ], "metadata": { "id": "Csg-bpJuoIXK" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "2poXnlF5XEXN" } }, { "cell_type": "markdown", "source": [ "## Part 7: Regression-to-Classification" ], "metadata": { "id": "-e5pgBLLoN81" } }, { "cell_type": "markdown", "source": [ "In this section, you will **reframe your original regression problem as a classification problem**.\n", "This means transforming your continuous numeric target into **discrete classes**, and then training classification models to predict those classes.\n", "\n", "\n", "\n", "\n", "\n" ], "metadata": { "id": "Qtj_Sf-7oqY5" } }, { "cell_type": "markdown", "source": [ "\n", "#### **7.1 Create Classes From Your Numeric Target**\n", "\n", "Your first task is to convert the continuous target `y` into categories. Choose a strategy to convert your numeric target into classes. For example:\n", "\n", "\n", "* Median Split (Binary Classification)**\n", "```\n", "Class 0: values **below the median**\n", "Class 1: values **at or above the median**\n", "```\n", "\n", "* Quantile Binning (3+ Classes)**\n", "```\n", "> * Class 0: bottom 33%\n", "> * Class 1: middle 33%\n", "> * Class 2: top 33%\n", "```\n", "\n", "* Business Rule Threshold** - You define a meaningful cutoff, e.g.:\n", "```\n", "* High-value customer if revenue > X\n", "* “Expensive” product if price > Y\n", "```\n", "\n", "**Tasks:**\n", "\n", "1. Implement your chosen strategy on the **train** and **test** targets. Using the **same engineered features** as before.\n", "\n", "2. Explain the reasoning behind your choice (2–3 sentences)." ], "metadata": { "id": "M9g9bfxlqWYg" } }, { "cell_type": "markdown", "source": [ "### 7.1 Strategy : Business Rule Threshold\n", "\n", "We convert `Benefit per order` into two classes based on whether the order is profitable or not. Class 0 represents a loss (profit below 0 USD) and Class 1 represents a profitable order (profit at or above 0 USD).\n", "\n", "This threshold is the most meaningful business cutoff in the dataset. Knowing in advance whether an order will result in a loss allows the business to intervene (adjusting pricing, flagging the order for review, or renegotiating shipping terms before it is too late)." ], "metadata": { "id": "VNRTB0uJjqkb" } }, { "cell_type": "code", "source": [ "# Step 1 : Convert regression target into binary classes\n", "# Class 0 = loss order (profit < 0)\n", "# Class 1 = profitable order (profit >= 0)\n", "\n", "y_class_train = (y_train_eng >= 0).astype(int)\n", "y_class_test = (y_test_eng >= 0).astype(int)\n", "\n", "print(\"Train class distribution:\")\n", "print(y_class_train.value_counts(normalize=True).round(3))\n", "print(\"\\nTest class distribution:\")\n", "print(y_class_test.value_counts(normalize=True).round(3))" ], "metadata": { "id": "IqC0YSXxqL3E", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "221d236b-564d-41d3-a526-fdac5813b372" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train class distribution:\n", "Benefit per order\n", "1 0.812\n", "0 0.188\n", "Name: proportion, dtype: float64\n", "\n", "Test class distribution:\n", "Benefit per order\n", "1 0.815\n", "0 0.185\n", "Name: proportion, dtype: float64\n" ] } ] }, { "cell_type": "code", "source": [ "# Step 2 : Visualize class balance\n", "class_counts = y_class_train.value_counts().reset_index()\n", "class_counts.columns = ['Class', 'Count']\n", "class_counts['Class'] = class_counts['Class'].map({0: 'Loss (Class 0)', 1: 'Profitable (Class 1)'})\n", "\n", "fig = px.bar(\n", " class_counts,\n", " x='Class', y='Count',\n", " title='Class Distribution : Loss vs Profitable Orders',\n", " color='Class',\n", " text='Count',\n", " color_discrete_sequence=['#e07070', '#70b8a0']\n", ")\n", "fig.update_traces(texttemplate='%{text:,}', textposition='outside')\n", "fig.update_layout(height=420, showlegend=False)\n", "fig.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 437 }, "id": "EsYjFM16jxSo", "outputId": "a9f9bfda-ebaa-4b87-f168-9be279450570" }, "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", "\n", "\n", "
\n", "
\n", "\n", "" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "\n", "#### **7.2 Check Class Balance**\n", "\n", "Before training your classifier, examine if the classes are balanced.\n", "\n", "1. Show the resulting **class distribution** (counts or percentages).\n", "2. Are some classes under-represented?\n", "3. If the data is imbalanced, explain which metric you’ll focus on (e.g., F1 score, recall) and why accuracy alone is misleading.\n", "4. If needed, consider changing your convertion." ], "metadata": { "id": "L8Grz1xfqLfH" } }, { "cell_type": "markdown", "source": [ "### 7.2 Class Balance\n", "\n", "The dataset is imbalanced: 81.2% of orders are profitable (Class 1) and only 18.8% are losses (Class 0). The split is almost identical between train and test, which confirms the split was clean.\n", "\n", "Because the classes are imbalanced, accuracy alone is misleading as a metric. A model that predicts \"profitable\" for every single order would score 81% accuracy without learning anything useful. Instead we will focus on **F1 score** and **recall for Class 0**, since correctly identifying loss orders is the business-critical task. Missing a loss order (false negative) is more costly than incorrectly flagging a profitable one (false positive)." ], "metadata": { "id": "UAzJ-cdSk_hH" } }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "75xPnIjfqHEh" } }, { "cell_type": "markdown", "source": [ "# Part 8: Train & Eval Classification Models\n", "\n" ], "metadata": { "id": "Ii0otL-qqHOt" } }, { "cell_type": "markdown", "source": [ "#### 8.1 Answer the following here, and later mention it in your presentation.\n", "\n" ], "metadata": { "id": "co-38v1UsPd2" } }, { "cell_type": "markdown", "source": [ "In the context of your dataset/task, explain what would be more importatnt - precision or recall.\n", "\n" ], "metadata": { "id": "1XM_xVUGsfon" } }, { "cell_type": "markdown", "source": [ "In the context of your dataset/task, explain what would be more critical - False Positive or False Negative.\n" ], "metadata": { "id": "xbrqOTcBsf2Q" } }, { "cell_type": "markdown", "source": [ "# Precision vs Recall:\n", "# Recall is more important in our context.\n", "# We are trying to catch loss orders before they happen.\n", "# Missing a loss order (false negative) costs the business real money.\n", "# It is better to flag too many orders for review than to miss a loss entirely.\n", "\n", "# False Positive vs False Negative:\n", "# A False Negative is more critical here.\n", "# If the model predicts an order will be profitable but it actually loses money,\n", "# the business takes a direct financial hit with no chance to intervene.\n", "# A False Positive just means an order gets flagged unnecessarily, which is a minor inconvenience." ], "metadata": { "id": "TJ1E876TlktE" } }, { "cell_type": "markdown", "source": [ "#### 8.2: Train **three** different kinds of classification models.\n", "\n", "\n", "Go to SKlearn to find different classification models. And use them." ], "metadata": { "id": "8nZOjcjZsocr" } }, { "cell_type": "code", "source": [ "# Step 1 : Train three different classification models on the same engineered features\n", "# Logistic Regression : linear classifier, fast and interpretable\n", "# Random Forest : ensemble of trees, handles non-linear patterns\n", "# Gradient Boosting : sequential ensemble, typically strong on tabular data\n", "\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n", "\n", "lr_clf = LogisticRegression(max_iter=1000, random_state=SEED)\n", "lr_clf.fit(X_train_eng, y_class_train)\n", "\n", "rf_clf = RandomForestClassifier(n_estimators=100, random_state=SEED, n_jobs=-1)\n", "rf_clf.fit(X_train_eng, y_class_train)\n", "\n", "gb_clf = GradientBoostingClassifier(n_estimators=100, random_state=SEED)\n", "gb_clf.fit(X_train_eng, y_class_train)\n", "\n", "print(\"All three models trained.\")" ], "metadata": { "id": "6MIie1Rws3pE", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "b3f41043-a2f2-4a50-987c-0b13c4f7f68d" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "All three models trained.\n" ] } ] }, { "cell_type": "markdown", "source": [ "#### 8.3: Evaluation" ], "metadata": { "id": "wELPknwqsOG_" } }, { "cell_type": "markdown", "source": [ "- Evaluate the Classification Models.\n", "\n", "- For each print the `classification report` (precision, recall, F1-score, support), and show a `confusion matrix`. (use SKlean built tools) Comment on what types of mistakes the model makes (based on the confusion matrix).\n", "\n", "- Identify which model performs best and try explain **why**.\n", "\n" ], "metadata": { "id": "bEzxJLmVsvVx" } }, { "cell_type": "code", "source": [ "# Step 2 : Evaluate all three models\n", "# classification_report shows precision, recall, F1 and support for each class\n", "# confusion matrix shows exactly what types of mistakes each model makes\n", "\n", "from sklearn.metrics import classification_report, confusion_matrix, ConfusionMatrixDisplay\n", "import matplotlib.pyplot as plt\n", "\n", "models = {\n", " 'Logistic Regression': lr_clf,\n", " 'Random Forest': rf_clf,\n", " 'Gradient Boosting': gb_clf\n", "}\n", "\n", "for name, model in models.items():\n", " y_pred = model.predict(X_test_eng)\n", " print(f\"\\n{'='*50}\")\n", " print(f\"{name}\")\n", " print('='*50)\n", " print(classification_report(y_class_test, y_pred, target_names=['Loss', 'Profitable']))\n", "\n", " cm = confusion_matrix(y_class_test, y_pred)\n", " disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=['Loss', 'Profitable'])\n", " fig, ax = plt.subplots(figsize=(5, 4))\n", " disp.plot(ax=ax, colorbar=False, cmap='Blues')\n", " ax.set_title(f'Confusion Matrix : {name}')\n", " plt.tight_layout()\n", " plt.show()" ], "metadata": { "id": "Rozrn2petFKA", "colab": { "base_uri": "https://localhost:8080/", "height": 1000 }, "outputId": "26ab41f5-c833-4d23-d495-d7797a4f6434" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\n", "==================================================\n", "Logistic Regression\n", "==================================================\n", " precision recall f1-score support\n", "\n", " Loss 0.88 0.05 0.10 6664\n", " Profitable 0.82 1.00 0.90 29440\n", "\n", " accuracy 0.82 36104\n", " macro avg 0.85 0.53 0.50 36104\n", "weighted avg 0.83 0.82 0.75 36104\n", "\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": { "image/png": { "width": 425, "height": 388 } } }, { "output_type": "stream", "name": "stdout", "text": [ "\n", "==================================================\n", "Random Forest\n", "==================================================\n", " precision recall f1-score support\n", "\n", " Loss 0.58 0.04 0.08 6664\n", " Profitable 0.82 0.99 0.90 29440\n", "\n", " accuracy 0.82 36104\n", " macro avg 0.70 0.52 0.49 36104\n", "weighted avg 0.78 0.82 0.75 36104\n", "\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": { "image/png": { "width": 419, "height": 388 } } }, { "output_type": "stream", "name": "stdout", "text": [ "\n", "==================================================\n", "Gradient Boosting\n", "==================================================\n", " precision recall f1-score support\n", "\n", " Loss 1.00 0.01 0.01 6664\n", " Profitable 0.82 1.00 0.90 29440\n", "\n", " accuracy 0.82 36104\n", " macro avg 0.91 0.50 0.46 36104\n", "weighted avg 0.85 0.82 0.74 36104\n", "\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": { "image/png": { "width": 421, "height": 388 } } } ] }, { "cell_type": "markdown", "source": [ "## 8.3 Model Comparison & Evaluation\n", "\n", "All three models hit about 82% accuracy, but accuracy is misleading here because 82% of the dataset is Profitable. A model that predicts \"Profitable\" every single time would also score 82%. The real test is whether the model can catch Loss orders.\n", "\n", "| Model | Loss Recall | Loss F1 | Macro F1 |\n", "|---|---|---|---|\n", "| Logistic Regression | 0.05 | 0.10 | 0.50 |\n", "| Random Forest | 0.04 | 0.08 | 0.49 |\n", "| Gradient Boosting | 0.01 | 0.01 | 0.46 |\n", "\n", "Logistic Regression comes out on top, which is a bit surprising given that Random Forest won the regression task. It catches 5% of loss orders versus 4% for RF and barely 1% for GB. Gradient Boosting went to the other extreme: when it did predict Loss, it was always right (precision 1.00), but it only made that call 74 times out of 6,664 actual losses, making it effectively useless for this task.\n", "\n", "The core issue across all three models is class imbalance: only 18.5% of orders are losses. The models learned that predicting Profitable almost always is a safe bet. To genuinely fix this, techniques like class_weight='balanced' or oversampling (SMOTE) would be needed, but that is beyond the scope of this assignment." ], "metadata": { "id": "0foKVNjCoZqx" } }, { "cell_type": "markdown", "source": [ "#### 8.4: Winner" ], "metadata": { "id": "iOW61vwUtEi1" } }, { "cell_type": "markdown", "source": [ "- Choose the best one out of the three models.\n", "\n", "- Export the model to a `pickle` file.\n", "\n", "- Upload the pickle file to the same(!) model repository.\n" ], "metadata": { "id": "2t9YFU9KtH8p" } }, { "cell_type": "markdown", "source": [ "## 8.4 Winning Classifier\n", "\n", "Winner: Logistic Regression\n", "\n", "Chosen based on highest macro F1 (0.50) and best recall for Loss orders (0.05). It is not a strong classifier for the minority class, but it outperforms both tree-based models on the metric that matters most for this business problem. The simplest model won, a reminder that complexity does not guarantee better results, especially when the underlying data has a class imbalance problem that no model can fully overcome without additional techniques." ], "metadata": { "id": "Y6upyah9ot52" } }, { "cell_type": "code", "source": [ "# RUN ONCE ONLY : exports and downloads the regression model\n", "# skip this cell after the first run\n", "from google.colab import files\n", "\n", "with open('classification_model.pkl', 'wb') as f:\n", " pickle.dump(lr_clf, f)\n", "\n", "print(\"classification_model.pkl saved.\")\n", "files.download('classification_model.pkl')" ], "metadata": { "id": "S53cFO8RpF-W" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "source": [ "# ============================================================\n", "# Export all key charts as PNG: matplotlib only, no kaleido\n", "# ============================================================\n", "\n", "import os, zipfile\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "import matplotlib.ticker as mticker\n", "import seaborn as sns\n", "from scipy import stats\n", "from sklearn.cluster import KMeans\n", "from sklearn.decomposition import PCA\n", "from google.colab import files\n", "\n", "sns.set_theme(style='whitegrid', font_scale=1.1)\n", "os.makedirs('chart_exports', exist_ok=True)\n", "\n", "# ============================================================\n", "# PART 2 — Q1: Late delivery rate by shipping mode\n", "# ============================================================\n", "\n", "q1 = df.groupby('Shipping Mode')['Late_delivery_risk'].mean().sort_values(ascending=False) * 100\n", "\n", "fig, ax = plt.subplots(figsize=(8, 5))\n", "bars = ax.bar(q1.index, q1.values, color=sns.color_palette('Set2', len(q1)))\n", "ax.set_title('Late Delivery Rate by Shipping Mode', fontsize=14, fontweight='bold')\n", "ax.set_ylabel('Late Delivery Rate (%)')\n", "ax.set_ylim(0, 110)\n", "for bar, val in zip(bars, q1.values):\n", " ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 1.5,\n", " f'{val:.1f}%', ha='center', fontsize=11, fontweight='bold')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/q1_shipping_mode.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Q1 saved\")\n", "\n", "# ============================================================\n", "# PART 2 — Q2: Order volume and profit over time\n", "# ============================================================\n", "\n", "df['order_month_str'] = df['order date (DateOrders)'].dt.to_period('M').astype(str)\n", "q2 = df.groupby('order_month_str').agg(\n", " order_count=('Benefit per order', 'count'),\n", " avg_profit=('Benefit per order', 'mean')\n", ").reset_index()\n", "\n", "fig, ax1 = plt.subplots(figsize=(12, 5))\n", "ax2 = ax1.twinx()\n", "ax1.bar(q2['order_month_str'], q2['order_count'], color='steelblue', alpha=0.6, label='Order Count')\n", "ax2.plot(q2['order_month_str'], q2['avg_profit'], color='tomato', linewidth=2, marker='o', markersize=3, label='Avg Profit (USD)')\n", "ax1.set_title('Order Volume and Average Profit Over Time', fontsize=14, fontweight='bold')\n", "ax1.set_ylabel('Order Count', color='steelblue')\n", "ax2.set_ylabel('Avg Profit (USD)', color='tomato')\n", "ax1.tick_params(axis='x', rotation=45)\n", "ax1.set_xticks(ax1.get_xticks()[::3])\n", "fig.legend(loc='upper right', bbox_to_anchor=(0.9, 0.88))\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/q2_order_volume.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Q2 saved\")\n", "\n", "# ============================================================\n", "# PART 2 — Q3: World map saved as HTML (no kaleido needed)\n", "# ============================================================\n", "\n", "import plotly.express as px\n", "\n", "country_translations = {\n", " 'Estados Unidos': 'United States', 'México': 'Mexico', 'Francia': 'France',\n", " 'Alemania': 'Germany', 'Australia': 'Australia', 'Brasil': 'Brazil',\n", " 'Reino Unido': 'United Kingdom', 'Nigeria': 'Nigeria', 'China': 'China',\n", " 'Pakistán': 'Pakistan', 'India': 'India', 'Indonesia': 'Indonesia',\n", " 'Bangladesh': 'Bangladesh', 'Turquía': 'Turkey', 'Japón': 'Japan',\n", " 'Filipinas': 'Philippines', 'Etiopía': 'Ethiopia', 'Vietnam': 'Vietnam',\n", " 'Egipto': 'Egypt', 'Tailandia': 'Thailand', 'España': 'Spain',\n", " 'Italia': 'Italy', 'Canadá': 'Canada', 'Corea del Sur': 'South Korea',\n", " 'Colombia': 'Colombia', 'Argentina': 'Argentina', 'Polonia': 'Poland',\n", " 'Kenia': 'Kenya', 'Sudáfrica': 'South Africa', 'Arabia Saudita': 'Saudi Arabia',\n", " 'Irak': 'Iraq', 'Venezuela': 'Venezuela', 'Perú': 'Peru',\n", " 'Ucrania': 'Ukraine', 'Holanda': 'Netherlands', 'Malasia': 'Malaysia',\n", " 'Guatemala': 'Guatemala', 'Honduras': 'Honduras', 'Ecuador': 'Ecuador',\n", " 'Bolivia': 'Bolivia', 'Chile': 'Chile', 'Suecia': 'Sweden',\n", " 'Portugal': 'Portugal', 'Bélgica': 'Belgium', 'Austria': 'Austria',\n", " 'Grecia': 'Greece', 'Suiza': 'Switzerland', 'Dinamarca': 'Denmark',\n", " 'Noruega': 'Norway', 'Finlandia': 'Finland', 'Irlanda': 'Ireland',\n", " 'Nueva Zelanda': 'New Zealand', 'Irán': 'Iran', 'Argelia': 'Algeria',\n", " 'Emiratos Árabes Unidos': 'United Arab Emirates', 'Israel': 'Israel',\n", " 'Myanmar': 'Myanmar', 'Sri Lanka': 'Sri Lanka', 'Nepal': 'Nepal',\n", " 'Haití': 'Haiti', 'República Dominicana': 'Dominican Republic',\n", " 'Japón': 'Japan', 'Rumania': 'Romania', 'Kazajistán': 'Kazakhstan'\n", "}\n", "df['country_en'] = df['Order Country'].map(country_translations).fillna(df['Order Country'])\n", "q3 = df.groupby('country_en')['Late_delivery_risk'].mean().reset_index()\n", "q3.columns = ['Country', 'Late Delivery Risk']\n", "fig_q3 = px.choropleth(\n", " q3, locations='Country', locationmode='country names',\n", " color='Late Delivery Risk', color_continuous_scale='Reds',\n", " title='Average Late Delivery Risk by Country', range_color=[0, 1]\n", ")\n", "fig_q3.write_html('chart_exports/q3_world_map.html')\n", "print(\"Q3 saved as HTML — open it in browser, take a screenshot, save as q3_world_map.png\")\n", "\n", "# ============================================================\n", "# PART 2 — Q4: Discount vs profit scatter\n", "# ============================================================\n", "\n", "sample = df.sample(8000, random_state=42)\n", "colors = sample['Shipping Mode'].map({\n", " 'First Class': '#e74c3c', 'Second Class': '#3498db',\n", " 'Standard Class': '#2ecc71', 'Same Day': '#f39c12'\n", "})\n", "\n", "fig, ax = plt.subplots(figsize=(9, 5))\n", "ax.scatter(sample['Order Item Discount Rate'], sample['Benefit per order'],\n", " c=colors, alpha=0.3, s=8)\n", "ax.axhline(0, color='red', linestyle='--', linewidth=1.2, label='Break-even')\n", "ax.set_xlabel('Discount Rate')\n", "ax.set_ylabel('Profit per Order (USD)')\n", "ax.set_title('Discount Rate vs Profit per Order', fontsize=14, fontweight='bold')\n", "handles = [plt.Line2D([0],[0], marker='o', color='w', markerfacecolor=c, markersize=8, label=l)\n", " for l, c in {'First Class':'#e74c3c','Second Class':'#3498db',\n", " 'Standard Class':'#2ecc71','Same Day':'#f39c12'}.items()]\n", "ax.legend(handles=handles, title='Shipping Mode')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/q4_discount_profit.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Q4 saved\")\n", "\n", "# ============================================================\n", "# PART 2 — Q5: Profit distribution KDE\n", "# ============================================================\n", "\n", "profits = df['Benefit per order'].dropna().values\n", "kde = stats.gaussian_kde(profits)\n", "x_range = np.linspace(profits.min(), profits.max(), 500)\n", "y_kde = kde(x_range)\n", "\n", "fig, ax = plt.subplots(figsize=(10, 5))\n", "ax.fill_between(x_range, y_kde, where=(x_range < 0), alpha=0.5, color='tomato', label='Loss orders (18.7%)')\n", "ax.fill_between(x_range, y_kde, where=(x_range >= 0), alpha=0.5, color='steelblue', label='Profitable orders (81.3%)')\n", "ax.plot(x_range, y_kde, color='black', linewidth=1.5)\n", "ax.axvline(0, color='red', linestyle='--', linewidth=1.5, label='Break-even (0 USD)')\n", "ax.set_xlabel('Benefit per Order (USD)')\n", "ax.set_ylabel('Density')\n", "ax.set_title('Profit Distribution: Loss vs Profitable Orders', fontsize=14, fontweight='bold')\n", "ax.legend()\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/q5_profit_distribution.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Q5 saved\")\n", "\n", "# ============================================================\n", "# PART 2 — Q6: Shipping schedule adherence violin\n", "# ============================================================\n", "\n", "df['shipping_gap'] = df['Days for shipping (real)'] - df['Days for shipment (scheduled)']\n", "\n", "fig, ax = plt.subplots(figsize=(9, 5))\n", "modes = df['Shipping Mode'].unique()\n", "data_by_mode = [df[df['Shipping Mode'] == m]['shipping_gap'].dropna().values for m in modes]\n", "parts = ax.violinplot(data_by_mode, showmedians=True, showextrema=True)\n", "for pc, color in zip(parts['bodies'], sns.color_palette('Set2', len(modes))):\n", " pc.set_facecolor(color)\n", " pc.set_alpha(0.7)\n", "ax.axhline(0, color='red', linestyle='--', linewidth=1.2, label='On time')\n", "ax.set_xticks(range(1, len(modes)+1))\n", "ax.set_xticklabels(modes)\n", "ax.set_ylabel('Days Late (positive = late)')\n", "ax.set_title('Shipping Schedule Adherence by Mode', fontsize=14, fontweight='bold')\n", "ax.legend()\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/q6_violin.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Q6 saved\")\n", "\n", "# ============================================================\n", "# PART 3 — Baseline: predicted vs actual\n", "# ============================================================\n", "\n", "y_pred_baseline = baseline_model.predict(X_test)\n", "fig, ax = plt.subplots(figsize=(7, 5))\n", "ax.scatter(y_test, y_pred_baseline, alpha=0.15, s=5, color='steelblue')\n", "ax.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()],\n", " 'r--', lw=1.5, label='Perfect prediction')\n", "ax.set_xlabel('Actual Profit (USD)')\n", "ax.set_ylabel('Predicted Profit (USD)')\n", "ax.set_title('Baseline Linear Regression: Predicted vs Actual', fontsize=13, fontweight='bold')\n", "ax.legend()\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part3_predicted_vs_actual.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 3 saved\")\n", "\n", "# ============================================================\n", "# PART 4 — Elbow plot\n", "# ============================================================\n", "\n", "inertias = []\n", "for k in range(1, 11):\n", " km = KMeans(n_clusters=k, random_state=SEED, n_init=10)\n", " km.fit(df_eng_scaled)\n", " inertias.append(km.inertia_)\n", "\n", "fig, ax = plt.subplots(figsize=(8, 5))\n", "ax.plot(range(1, 11), inertias, marker='o', color='steelblue', linewidth=2)\n", "ax.axvline(4, color='red', linestyle='--', linewidth=1.2, label='k=4 chosen')\n", "ax.set_xlabel('Number of Clusters (k)')\n", "ax.set_ylabel('Inertia')\n", "ax.set_title('KMeans Elbow Plot', fontsize=14, fontweight='bold')\n", "ax.legend()\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part4_elbow.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 4 elbow saved\")\n", "\n", "# ============================================================\n", "# PART 4 — PCA cluster scatter\n", "# ============================================================\n", "\n", "pca_vis = PCA(n_components=2, random_state=SEED)\n", "sample_idx = df_eng_scaled.sample(10000, random_state=SEED).index\n", "components = pca_vis.fit_transform(df_eng_scaled.loc[sample_idx])\n", "cluster_labels = df.loc[sample_idx, 'cluster'].values\n", "\n", "fig, ax = plt.subplots(figsize=(8, 6))\n", "palette = sns.color_palette('Set1', 4)\n", "for c in sorted(np.unique(cluster_labels)):\n", " mask = cluster_labels == c\n", " ax.scatter(components[mask, 0], components[mask, 1],\n", " s=5, alpha=0.4, color=palette[c], label=f'Cluster {c}')\n", "ax.set_xlabel('PC1')\n", "ax.set_ylabel('PC2')\n", "ax.set_title('KMeans Clusters Visualized with PCA (10k sample)', fontsize=13, fontweight='bold')\n", "ax.legend(markerscale=3)\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part4_pca_clusters.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 4 PCA saved\")\n", "\n", "# ============================================================\n", "# PART 5 — Model comparison bar chart\n", "# ============================================================\n", "\n", "models_names = ['Baseline LR', 'LR (engineered)', 'Random Forest', 'Gradient Boosting']\n", "r2_scores = [0.016, 0.318, 0.481, 0.198]\n", "colors_comp = ['#95a5a6', '#3498db', '#2ecc71', '#e67e22']\n", "\n", "fig, ax = plt.subplots(figsize=(8, 5))\n", "bars = ax.bar(models_names, r2_scores, color=colors_comp)\n", "ax.set_ylabel('R2 Score')\n", "ax.set_title('Regression Model Comparison: R2 Score', fontsize=14, fontweight='bold')\n", "ax.set_ylim(0, 0.6)\n", "for bar, val in zip(bars, r2_scores):\n", " ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 0.01,\n", " f'{val:.3f}', ha='center', fontsize=11, fontweight='bold')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part5_model_comparison.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 5 comparison saved\")\n", "\n", "# ============================================================\n", "# PART 5 — Random Forest feature importance\n", "# ============================================================\n", "\n", "rf_imp = pd.DataFrame({'Feature': X_train_eng.columns, 'Importance': rf.feature_importances_})\n", "rf_imp = rf_imp.sort_values('Importance').tail(15)\n", "\n", "fig, ax = plt.subplots(figsize=(9, 6))\n", "bars = ax.barh(rf_imp['Feature'], rf_imp['Importance'],\n", " color=sns.color_palette('Blues_d', len(rf_imp)))\n", "ax.set_xlabel('Importance')\n", "ax.set_title('Random Forest: Top 15 Feature Importances', fontsize=13, fontweight='bold')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part5_rf_importance.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 5 RF importance saved\")\n", "\n", "# ============================================================\n", "# PART 5 — Gradient Boosting feature importance\n", "# ============================================================\n", "\n", "gb_imp = pd.DataFrame({'Feature': X_train_eng.columns, 'Importance': gb.feature_importances_})\n", "gb_imp = gb_imp.sort_values('Importance').tail(15)\n", "\n", "fig, ax = plt.subplots(figsize=(9, 6))\n", "ax.barh(gb_imp['Feature'], gb_imp['Importance'],\n", " color=sns.color_palette('Oranges_d', len(gb_imp)))\n", "ax.set_xlabel('Importance')\n", "ax.set_title('Gradient Boosting: Top 15 Feature Importances', fontsize=13, fontweight='bold')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part5_gb_importance.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 5 GB importance saved\")\n", "\n", "# ============================================================\n", "# PART 7 — Class distribution\n", "# ============================================================\n", "\n", "class_labels = ['Profitable', 'Loss']\n", "class_counts = [(y_class_train == 1).sum(), (y_class_train == 0).sum()]\n", "class_pcts = [81.2, 18.8]\n", "\n", "fig, ax = plt.subplots(figsize=(6, 5))\n", "bars = ax.bar(class_labels, class_counts, color=['steelblue', 'tomato'])\n", "ax.set_ylabel('Number of Orders')\n", "ax.set_title('Class Distribution: Loss vs Profitable Orders', fontsize=13, fontweight='bold')\n", "for bar, pct in zip(bars, class_pcts):\n", " ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 200,\n", " f'{pct}%', ha='center', fontsize=12, fontweight='bold')\n", "plt.tight_layout()\n", "plt.savefig('chart_exports/part7_class_distribution.png', dpi=150, bbox_inches='tight')\n", "plt.close()\n", "print(\"Part 7 saved\")\n", "\n", "# ============================================================\n", "# ZIP AND DOWNLOAD\n", "# ============================================================\n", "\n", "with zipfile.ZipFile('charts.zip', 'w') as zf:\n", " for fname in sorted(os.listdir('chart_exports')):\n", " zf.write(f'chart_exports/{fname}', fname)\n", "\n", "print(\"\\nAll done. Files exported:\")\n", "for fname in sorted(os.listdir('chart_exports')):\n", " print(f\" {fname}\")\n", "\n", "files.download('charts.zip')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 513 }, "id": "K_oQVxrk6tcy", "outputId": "170d6cdb-f586-489c-9134-b42a06be2af0" }, "execution_count": null, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Q1 saved\n", "Q2 saved\n", "Q3 saved as HTML — open it in browser, take a screenshot, save as q3_world_map.png\n", "Q4 saved\n", "Q5 saved\n", "Q6 saved\n", "Part 3 saved\n", "Part 4 elbow saved\n", "Part 4 PCA saved\n", "Part 5 comparison saved\n", "Part 5 RF importance saved\n", "Part 5 GB importance saved\n", "Part 7 saved\n", "\n", "All done. Files exported:\n", " part3_predicted_vs_actual.png\n", " part4_elbow.png\n", " part4_pca_clusters.png\n", " part5_gb_importance.png\n", " part5_model_comparison.png\n", " part5_rf_importance.png\n", " part7_class_distribution.png\n", " q1_shipping_mode.png\n", " q2_order_volume.png\n", " q3_world_map.html\n", " q4_discount_profit.png\n", " q5_profit_distribution.png\n", " q6_violin.png\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "" ], "application/javascript": [ "\n", " async function download(id, filename, size) {\n", " if (!google.colab.kernel.accessAllowed) {\n", " return;\n", " }\n", " const div = document.createElement('div');\n", " const label = document.createElement('label');\n", " label.textContent = `Downloading \"${filename}\": `;\n", " div.appendChild(label);\n", " const progress = document.createElement('progress');\n", " progress.max = size;\n", " div.appendChild(progress);\n", " document.body.appendChild(div);\n", "\n", " const buffers = [];\n", " let downloaded = 0;\n", "\n", " const channel = await google.colab.kernel.comms.open(id);\n", " // Send a message to notify the kernel that we're ready.\n", " channel.send({})\n", "\n", " for await (const message of channel.messages) {\n", " // Send a message to notify the kernel that we're ready.\n", " channel.send({})\n", " if (message.buffers) {\n", " for (const buffer of message.buffers) {\n", " buffers.push(buffer);\n", " downloaded += buffer.byteLength;\n", " progress.value = downloaded;\n", " }\n", " }\n", " }\n", " const blob = new Blob(buffers, {type: 'application/binary'});\n", " const a = document.createElement('a');\n", " a.href = window.URL.createObjectURL(blob);\n", " a.download = filename;\n", " div.appendChild(a);\n", " a.click();\n", " div.remove();\n", " }\n", " " ] }, "metadata": {} }, { "output_type": "display_data", "data": { "text/plain": [ "" ], "application/javascript": [ "download(\"download_ba1a53b7-2f5a-4015-8ff3-11fd6e5ce254\", \"charts.zip\", 5654369)" ] }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "TlClTFEgXFTX" } }, { "cell_type": "markdown", "source": [ "# **Part 9: Presentation Video**\n", "\n", "- Record a brief video (4–6 minutes) with screen sharing of you walk through the HF's model repository, README, and sharing your process & results. Include a screen share while also recording yourself talking during the walk through.\n", "\n", "- The recording will include sharing the screen, and you talking to the camera (show yourself in a circle on the bottom).\n", "\n", "- Videos without your face talking while going ower your work wont be acceptable.\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n" ], "metadata": { "id": "mzYCO54GIPdP" } }, { "cell_type": "markdown", "source": [ "- Include:\n", " - A quick dataset overview and your main goal.\n", " - Key EDA steps and highlights of visual insights.\n", " - How you engineered features. About your clustering.\n", " - The models you trained, your iterative process, and what you learned.\n", " - Key visualizations and takeaways.\n", " - Reflections on any challenges and lessons learned.\n", " - Extra work.\n", "\n", "- Finally, attach the video to the beginning of the `README` file, and make sure everything works. *The video should be placed at the beginning of the README and must be playable within it. It can be recorded using `Vimeo`, `YouTube`, `Loom`, and uploaded to your HF model repo.*" ], "metadata": { "id": "Kw06OJESuWGp" } }, { "cell_type": "code", "source": [ "# The following is an example on how to include your video in the README file.\n", "# \n" ], "metadata": { "id": "ccJCoq6HutHy" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "


\n", "\n", "---\n", "\n", "


" ], "metadata": { "id": "YvNRPdxhvWaK" } }, { "cell_type": "markdown", "source": [ "# Part 10: Moodle" ], "metadata": { "id": "EZzOzA3YupFc" } }, { "cell_type": "markdown", "source": [ "**Submit to Moodle only one link - the link to your HF's Model Repository.** \n", "\n", "The Repo should Include:\n", "- README\n", "- Python Notebook\n", "- 1 pickle model for regression\n", "- 1 pickle model for classification\n", "- Video Presentation" ], "metadata": { "id": "n28Xs2gRusQh" } }, { "cell_type": "code", "source": [], "metadata": { "id": "DFqtRAswvKox" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "\n", "---\n", "\n", "
\n", "
\n", "
\n", "
\n", "\n", "Good luck and have fun creating AI model!" ], "metadata": { "id": "rN8TJ_5oIPfm" } }, { "cell_type": "code", "source": [], "metadata": { "id": "cPgKfBWKvU8K" }, "execution_count": null, "outputs": [] } ] }