prjct3 customer.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"private_outputs":true,"provenance":[],"mount_file_id":"1VQNz6CKILYXjwC9Ae4CH8Xe4rJaqxa-y","authorship_tag":"ABX9TyOl/SAW6nxkduTELwRP+K8G"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"code","execution_count":null,"metadata":{"id":"f2buhy6wYCUM"},"outputs":[],"source":["#Importing all necessary libraries\n","import numpy as np\n","import pandas as pd\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","import warnings\n","warnings.filterwarnings('ignore')"]},{"cell_type":"code","source":["data = pd.read_csv('/content/drive/MyDrive/cognorise/Data analytics/8_Customertravel.csv')\n","data"],"metadata":{"id":"Fs9prBU9Yfc6"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["print(data)\n","\n","print('dimensions:')\n","print(data.shape)\n","\n","print('Information:')\n","data.info()"],"metadata":{"id":"EWzIwc1lnyxS"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["data.isnull().values.any()\n","value = len(data[data.duplicated()])\n","print(value)"],"metadata":{"id":"3yG5Zs9Wn-JP"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["print(\"Statistical Summary\")\n","data.describe().T"],"metadata":{"id":"e20OvtEgoD78"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["data.describe(include = 'object')"],"metadata":{"id":"WVcSgZvXoLkG"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["for col in data.describe(include = 'object') .columns:\n","    print(col)\n","    print(data[col].unique())\n","    print('-'*50)"],"metadata":{"id":"X5d5s3RooO3J"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Convert categorical columns to numerical using one-hot encoding\n","data_encoded = pd.get_dummies(data)\n","\n","# Calculate correlations on the encoded DataFrame\n","data_encoded.corr()"],"metadata":{"id":"hlnxZwJgoSKA"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["data.isnull().sum()"],"metadata":{"id":"t4a1xUvooVjP"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":[" Data Visualization"],"metadata":{"id":"WzffEPwUoxG2"}},{"cell_type":"code","source":["from sklearn.preprocessing import LabelEncoder\n","number = LabelEncoder()\n","features = ['FrequentFlyer','AnnualIncomeClass','AccountSyncedToSocialMedia','BookedHotelOrNot']\n","for f in features:\n","    data[f] = number.fit_transform(data[f])\n","data.head()"],"metadata":{"id":"5uwqJZ8Jotcj"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Hotel Booking Visualization"],"metadata":{"id":"jvVZ3x1jo6Dm"}},{"cell_type":"code","source":["churn_perc = data ['BookedHotelOrNot'].value_counts(normalize = 1)\n","print(churn_perc)\n","\n","plt.figure(figsize=(5,6))\n","plt.bar([\"Booked\",\"NotBooked\"], data ['BookedHotelOrNot'].value_counts(normalize = 1), edgecolor = 'black', width = 0.8, color = ['skyblue', 'purple'])\n","plt.title('How Hotel Bookings Affect Customer Churn')\n","plt.show()"],"metadata":{"id":"gm4RQME5o1yL"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Annual Income Visualization"],"metadata":{"id":"TggjCnT1pGG1"}},{"cell_type":"code","source":["income = data.groupby(\"AnnualIncomeClass\")\n","income = income.size()\n","income"],"metadata":{"id":"4R3toWK1o8bT"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["plt.pie(income.values , labels = (\"High Income\", \"Low Income\", \"Middle Income\" ), autopct='%1.1f%%', radius = 1.2, textprops = {\"fontsize\" : 16})\n","plt.title(\"How Income Impacts Customer Churn\", c=\"b\")\n","plt.show()\n"],"metadata":{"id":"BqbKvA1YpRWP"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Analytical Summary"],"metadata":{"id":"fSBhh0H9pZUP"}},{"cell_type":"markdown","source":["\n","\n","Customers who have made hotel reservations account for 60% of the overall population, whereas those who have not made hotel reservations account for 39% of the population, which is a significant number.\n","\n","We found that customers with the highest yearly income have the lowest proportion of customer churn, whereas those with the lowest (40%) and middle (42%), have the highest percentage of churn.\n"],"metadata":{"id":"E48rs_Ilpmv9"}},{"cell_type":"code","source":[],"metadata":{"id":"eveYk4fepUTJ"},"execution_count":null,"outputs":[]}]}