{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# H&M Sales and Customers analysis","metadata":{}},{"cell_type":"markdown","source":"The following project aims to analyze the sales and customers data provided by H&M.
\nThe project is structred as follows:\n- Transactions analysis:\n - Q1 - Which are the TOP 100 articles in terms of sold quantity?\n - Q2 - Are there articles that have been sold only once?\n - Q3 - Which are the TOP 100 articles that generated most earnings for the company?\n - Q4 - Which are articles that generated lower earnings for the company? \n\n- Customer Analysis:\n - Q5 - Which age group purchase more articles?\n - Q6 - Which age group generates more earnings for the company?\n - Q7 - Do active customers on the fashion news purchase more articles?\n - Q8 - Does the club member status influence the purchased quantity of articles?","metadata":{}},{"cell_type":"markdown","source":"# **Main results summary dashboard:**","metadata":{}},{"cell_type":"markdown","source":"![](\"https://i.imgur.com/8sn0J8D.png\")
","metadata":{}},{"cell_type":"markdown","source":"**Among the TOP 100 of solds products:**\n- Almost 30% of sold products are trousers\n- 38% is Ladieswear\n- 30% is Lingeries/Tights\n- Over 70% are black colored\n- Almost 40% are related to lower body","metadata":{}},{"cell_type":"markdown","source":"![](\"https://i.imgur.com/TG95am4.png\")
","metadata":{}},{"cell_type":"markdown","source":"**Among the TOP 100 products in terms of earnings:**\n- Over 60% of the total earnings are generated by selling trousers\n- Around 50% of these products belonsg to the DIVIDED collection (a H&M teenage collection)\n- 37% of the products are from the Ladieswear line\n- 55% of the products are black\n- 66.2% of the products are related to lower body\n","metadata":{}},{"cell_type":"markdown","source":"![](\"https://i.imgur.com/5SZMW8h.png\")
","metadata":{}},{"cell_type":"markdown","source":"Insights:\n- Customers in the range 20-30 are responsible for more than 42% of the total purchased products.\n- Customers in the range 16-20. 60-70 and 70+ are responsible for the 8% of the total purchased products\n- Customers in the range 30-40, 40-50 and 50-60 are responsible for 16% of purchased quantity each.\n\n**Indeed a very similar situation to the purchases quantity can be found in the earnings analysis, since customers who buys more, on average leads to higher earnings for the company.
\nThe age group 20-30 is by far responsible for the highest earnings for the company (41.9% of total earnings).**","metadata":{}},{"cell_type":"markdown","source":"![](\"https://i.imgur.com/TSALUiK.png\")
","metadata":{}},{"cell_type":"markdown","source":"We can see that **customers in the range 20-30 and 30-40 have the lowest percentage of fashion news frequency, while being the groups which buy the most**.
\nMoreover, **the frequency of customer that regulary check fashion news starts increasing from the range 40-50, with a peak value of 43.7% of regular/active users for customers in the range 70+ years old**. This means that **checking fashion news seems to be more effective for older customers, who still represent a small percentage of total sold products, while younger customers do not need to check the news to buy new products**.
\n**It could be effective for the company to invite younger customers (range 20-40) to check the news more frequently in order to increase the sold items.**","metadata":{}},{"cell_type":"markdown","source":"![](\"https://i.imgur.com/ifAvKwO.png\")
","metadata":{}},{"cell_type":"markdown","source":"**This plots shows that the average purchased quantity differs a lot among the categories**.
\nIn particular, **customers belonging to the ACTIVE clubs, purchase more products than other categories, while those in the \"pre-create\" category purchaes on average less than a third of third of active customers**.\n","metadata":{}},{"cell_type":"markdown","source":"### Import libraries","metadata":{}},{"cell_type":"code","source":"import numpy as np \nimport pandas as pd\nfrom pandasql import sqldf\n\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n\nplt.style.use('seaborn-white')\nsns.set_style(\"whitegrid\")\nsns.despine()\nplt.rc(\"figure\", autolayout=True)\nplt.rc(\"axes\", labelweight=\"bold\", labelsize=\"large\", titleweight=\"bold\", titlesize=14, titlepad=10)\n\nimport matplotlib as mpl\n\nmpl.rcParams['axes.spines.left'] = False\nmpl.rcParams['axes.spines.right'] = False\nmpl.rcParams['axes.spines.top'] = False\nmpl.rcParams['axes.spines.bottom'] = False\nplt.rcParams[\"font.weight\"] = \"bold\"\nplt.rcParams[\"axes.labelweight\"] = \"bold\"","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:16:51.466726Z","iopub.execute_input":"2022-02-26T14:16:51.467347Z","iopub.status.idle":"2022-02-26T14:16:52.683977Z","shell.execute_reply.started":"2022-02-26T14:16:51.467243Z","shell.execute_reply":"2022-02-26T14:16:52.683049Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Data import","metadata":{}},{"cell_type":"markdown","source":"First, we import the data.","metadata":{}},{"cell_type":"code","source":"df_a = pd.read_csv(\"/kaggle/input/h-and-m-personalized-fashion-recommendations/articles.csv\")\ndf_t = pd.read_csv(\"/kaggle/input/h-and-m-personalized-fashion-recommendations/transactions_train.csv\")\ndf_c = pd.read_csv(\"/kaggle/input/h-and-m-personalized-fashion-recommendations/customers.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:16:52.685421Z","iopub.execute_input":"2022-02-26T14:16:52.685659Z","iopub.status.idle":"2022-02-26T14:18:09.040827Z","shell.execute_reply.started":"2022-02-26T14:16:52.685631Z","shell.execute_reply":"2022-02-26T14:18:09.039815Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Articles dataframe analysis","metadata":{}},{"cell_type":"code","source":"df_a.head(1)","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:09.042191Z","iopub.execute_input":"2022-02-26T14:18:09.042459Z","iopub.status.idle":"2022-02-26T14:18:09.076544Z","shell.execute_reply.started":"2022-02-26T14:18:09.042426Z","shell.execute_reply":"2022-02-26T14:18:09.075950Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_a.columns","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:09.078275Z","iopub.execute_input":"2022-02-26T14:18:09.078987Z","iopub.status.idle":"2022-02-26T14:18:09.085440Z","shell.execute_reply.started":"2022-02-26T14:18:09.078944Z","shell.execute_reply":"2022-02-26T14:18:09.084415Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f\"The dataframe articles has {len(df_a)} rows\")","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:09.093422Z","iopub.execute_input":"2022-02-26T14:18:09.093714Z","iopub.status.idle":"2022-02-26T14:18:09.100775Z","shell.execute_reply.started":"2022-02-26T14:18:09.093677Z","shell.execute_reply":"2022-02-26T14:18:09.099848Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The \"articles\" dataframe has 25 columns and more than 100k rows.
\nFor our our analysis we will just select the following columns: \n- article_id\n- prod_name\n- product_type_name\n- product_group_name\n- colour_group_name\n- index_name\n\nBy considering only these columns we can also save lots of memory.","metadata":{}},{"cell_type":"markdown","source":"To filter the columns, we will use SQL like code through SQL-DF library.","metadata":{}},{"cell_type":"code","source":"df_a = sqldf(\"\"\"SELECT article_id, prod_name, product_type_name, product_group_name, colour_group_name, index_name\n FROM df_a\n \"\"\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:09.102415Z","iopub.execute_input":"2022-02-26T14:18:09.102779Z","iopub.status.idle":"2022-02-26T14:18:16.359888Z","shell.execute_reply.started":"2022-02-26T14:18:09.102743Z","shell.execute_reply":"2022-02-26T14:18:16.358662Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_a.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:16.361266Z","iopub.execute_input":"2022-02-26T14:18:16.361519Z","iopub.status.idle":"2022-02-26T14:18:16.375944Z","shell.execute_reply.started":"2022-02-26T14:18:16.361486Z","shell.execute_reply":"2022-02-26T14:18:16.374957Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Transactions dataframe analysis","metadata":{}},{"cell_type":"code","source":"df_t.head(1)","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:16.380426Z","iopub.execute_input":"2022-02-26T14:18:16.380803Z","iopub.status.idle":"2022-02-26T14:18:16.395337Z","shell.execute_reply.started":"2022-02-26T14:18:16.380752Z","shell.execute_reply":"2022-02-26T14:18:16.394554Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_t.columns","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:16.396246Z","iopub.execute_input":"2022-02-26T14:18:16.396507Z","iopub.status.idle":"2022-02-26T14:18:16.410713Z","shell.execute_reply.started":"2022-02-26T14:18:16.396478Z","shell.execute_reply":"2022-02-26T14:18:16.409632Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f\"The dataframe Transactions has {len(df_t)} rows\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:16.412824Z","iopub.execute_input":"2022-02-26T14:18:16.413861Z","iopub.status.idle":"2022-02-26T14:18:16.422507Z","shell.execute_reply.started":"2022-02-26T14:18:16.413810Z","shell.execute_reply":"2022-02-26T14:18:16.421531Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The Transactions dataframe has more than 31 million rows: in order to save memory, we decide to drop some columns and keep only \"customer_id\", \"article_id\", \"price\".","metadata":{}},{"cell_type":"code","source":"df_t = df_t[[\"customer_id\", \"article_id\", \"price\"]]","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:16.424369Z","iopub.execute_input":"2022-02-26T14:18:16.424813Z","iopub.status.idle":"2022-02-26T14:18:17.022354Z","shell.execute_reply.started":"2022-02-26T14:18:16.424673Z","shell.execute_reply":"2022-02-26T14:18:17.021668Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"In the following we will focus in the analysis of the transaction dataframe, in order to discover the most and least sold products.","metadata":{"execution":{"iopub.status.busy":"2022-02-20T17:46:34.27739Z","iopub.execute_input":"2022-02-20T17:46:34.277701Z","iopub.status.idle":"2022-02-20T17:46:34.28398Z","shell.execute_reply.started":"2022-02-20T17:46:34.277671Z","shell.execute_reply":"2022-02-20T17:46:34.282952Z"}}},{"cell_type":"markdown","source":"# Transactions analysis ","metadata":{}},{"cell_type":"markdown","source":"First, we extract the quantities sold per article using the value counts method on the \"article_id\" column of the transaction dataframe.","metadata":{}},{"cell_type":"code","source":"df_sold_qty = df_t[\"article_id\"].value_counts()\ndf_sold_qty","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:17.023430Z","iopub.execute_input":"2022-02-26T14:18:17.023796Z","iopub.status.idle":"2022-02-26T14:18:18.472638Z","shell.execute_reply.started":"2022-02-26T14:18:17.023767Z","shell.execute_reply":"2022-02-26T14:18:18.471740Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Then we create a dataframe based on this pandas series: this is necessary since later this dataframe will be joined with the \"article\" dataframe by the article_id column, in order to get informtions on the products.**","metadata":{}},{"cell_type":"code","source":"df_sold_qty=df_sold_qty.reset_index()\ndf_sold_qty.rename(columns = {\"article_id\":\"sold_qty\",\"index\":\"article_id\"}, inplace=True)\ndf_sold_qty.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:18.473865Z","iopub.execute_input":"2022-02-26T14:18:18.474137Z","iopub.status.idle":"2022-02-26T14:18:18.488695Z","shell.execute_reply.started":"2022-02-26T14:18:18.474106Z","shell.execute_reply":"2022-02-26T14:18:18.487750Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Then we can get some summary statistics about the sold quantities by calling the describe method on the \"sold_qty\" column:","metadata":{}},{"cell_type":"code","source":"df_sold_qty[\"sold_qty\"].describe()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:18.489981Z","iopub.execute_input":"2022-02-26T14:18:18.490365Z","iopub.status.idle":"2022-02-26T14:18:18.514995Z","shell.execute_reply.started":"2022-02-26T14:18:18.490330Z","shell.execute_reply":"2022-02-26T14:18:18.514285Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Summary statistics on the sold quantities:\n- there are 105000 different articles in the transactions.\n- There are items which have been sold only once\n- 25% of sold products, have been sold 14 or less times\n- 50% were sold 65 or less times\n- 75% were sold 286 or less times,\n- The most sold item have been sold 50287 times.","metadata":{}},{"cell_type":"markdown","source":"**We can expect a very skewed distribution of this variable, which can be checked by plotting the variable:**","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(12,4))\nplt.title(\"Sold Quantity KDE plot\")\nsns.kdeplot(df_sold_qty[\"sold_qty\"])\nplt.xlabel(\"Sold Quantity\")\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:18.516258Z","iopub.execute_input":"2022-02-26T14:18:18.516619Z","iopub.status.idle":"2022-02-26T14:18:19.304155Z","shell.execute_reply.started":"2022-02-26T14:18:18.516590Z","shell.execute_reply":"2022-02-26T14:18:19.303183Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Indeed, the distribution is heavily right skewed.**","metadata":{}},{"cell_type":"markdown","source":"It could be also interesting to check high quantiles of the distribution.","metadata":{}},{"cell_type":"code","source":"df_sold_qty[\"sold_qty\"].quantile([0.90,0.95,0.99,0.999])","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:19.305467Z","iopub.execute_input":"2022-02-26T14:18:19.306243Z","iopub.status.idle":"2022-02-26T14:18:19.318050Z","shell.execute_reply.started":"2022-02-26T14:18:19.306177Z","shell.execute_reply":"2022-02-26T14:18:19.317157Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The quantile analysis can give us the following insights: \n- 90% of the articles have been sold 793 or less times\n- 95% of the articles have been sold 1318 or less times\n- 99% of the articles have been sold 3185 or less times\n\n**This shows that there is a very small minority of items that sold more than 10k times (just the 0.001%), highlighting the skewness nature of the distribution.**","metadata":{}},{"cell_type":"markdown","source":"# Q1 - Which are the TOP 100 articles in terms of sold quantity?","metadata":{"execution":{"iopub.status.busy":"2022-02-21T08:43:45.895313Z","iopub.execute_input":"2022-02-21T08:43:45.895583Z","iopub.status.idle":"2022-02-21T08:43:45.913209Z","shell.execute_reply.started":"2022-02-21T08:43:45.895554Z","shell.execute_reply":"2022-02-21T08:43:45.912667Z"}}},{"cell_type":"markdown","source":"We can simply extract the most 100 sold items from the dataframe \"df_sold_qty\" by taking the first 100 rows.","metadata":{}},{"cell_type":"code","source":"top_100_sold = df_sold_qty.iloc[:100]\ntop_100_sold.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:19.319247Z","iopub.execute_input":"2022-02-26T14:18:19.319490Z","iopub.status.idle":"2022-02-26T14:18:19.334429Z","shell.execute_reply.started":"2022-02-26T14:18:19.319461Z","shell.execute_reply":"2022-02-26T14:18:19.333791Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Then we join this dataframe with the articles dataframe (df_a) by the \"article_id\" column in order to get more details about each article.","metadata":{}},{"cell_type":"code","source":"top_100_details = sqldf(\"\"\"SELECT *\n FROM top_100_sold t\n INNER JOIN df_a a\n on t.article_id = a.article_id\n \"\"\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:19.335621Z","iopub.execute_input":"2022-02-26T14:18:19.336196Z","iopub.status.idle":"2022-02-26T14:18:21.548617Z","shell.execute_reply.started":"2022-02-26T14:18:19.336153Z","shell.execute_reply":"2022-02-26T14:18:21.547620Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"top_100_details.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:21.550035Z","iopub.execute_input":"2022-02-26T14:18:21.550381Z","iopub.status.idle":"2022-02-26T14:18:21.565754Z","shell.execute_reply.started":"2022-02-26T14:18:21.550338Z","shell.execute_reply":"2022-02-26T14:18:21.564490Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(10,8))\nplt.title(\"TOP 30 most sold products\", fontsize=33, fontweight=\"bold\")\nno=30\ng = sns.barplot(y=\"prod_name\", x=\"sold_qty(%)\", data=top_100_details.iloc[:no].groupby(\"prod_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=12)\nplt.xlabel(\"Sold Quantity (%)\", size=25, fontweight=\"bold\")\nplt.ylabel(\"\")\nplt.grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:21.567452Z","iopub.execute_input":"2022-02-26T14:18:21.567974Z","iopub.status.idle":"2022-02-26T14:18:22.392086Z","shell.execute_reply.started":"2022-02-26T14:18:21.567921Z","shell.execute_reply":"2022-02-26T14:18:22.391045Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"We decided to plot only the TOP 30 of articles since including the TOP 50 or 100 product names would hve lead to a very big plot !\nIndeed we can see that among the TOP 100 sold articles:\n- The trousers \"Jade HW Skinny Denim TRS \" is responsible for almost 19% of all sold products.\n- the TOP 4 of most sold items, is responsible for almost 40% of the TOP 100 sold products.","metadata":{}},{"cell_type":"markdown","source":"For what concerns other product characteristics (besides the product name), we can obtain very effective plots even if we consider 100 products:","metadata":{}},{"cell_type":"code","source":"fig, ax = plt.subplots(2,2, figsize=(13,9.5))\nplt.suptitle(\"TOP 100 most sold products characteristics\", fontweight=\"bold\",fontsize=30)\n\nno=100\n\ng = sns.barplot(y=\"product_type_name\", x=\"sold_qty(%)\", data=top_100_details.iloc[:no].groupby(\"product_type_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[0,0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=12, color=\"black\")\nax[0,0].set_ylabel(\"\")\nax[0,0].set_xlabel(\"Sold Quantity (%)\", size=22, fontweight=\"bold\")\nax[0,0].set_title(\"Product Type\",fontweight=\"bold\",fontsize=28)\nax[0,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"index_name\", x=\"sold_qty(%)\", data=top_100_details.iloc[:no].groupby(\"index_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[0,1],palette=\"viridis\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15, color=\"black\")\nax[0,1].set_ylabel(\"\")\nax[0,1].set_xlabel(\"Sold Quantity (%)\", size=22, fontweight=\"bold\")\nax[0,1].set_title(\"Index\",fontweight=\"bold\",fontsize=28)\nax[0,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"colour_group_name\", x=\"sold_qty(%)\", data=top_100_details.iloc[:no].groupby(\"colour_group_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[1,0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15, color=\"black\")\nax[1,0].set_ylabel(\"\")\nax[1,0].set_xlabel(\"Sold Quantity (%)\", size=22, fontweight=\"bold\")\nax[1,0].set_title(\"Colour Group\",fontweight=\"bold\",fontsize=28)\nax[1,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"product_group_name\", x=\"sold_qty(%)\", data=top_100_details.iloc[:no].groupby(\"product_group_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[1,1],palette=\"Reds_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15, color=\"black\")\nax[1,1].set_ylabel(\"\")\nax[1,1].set_xlabel(\"Sold Quantity (%)\", size=22, fontweight=\"bold\")\nax[1,1].set_title(\"Product Group\",fontweight=\"bold\",fontsize=28)\nax[1,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\nfig.tight_layout()\n\nplt.show() ","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:22.393592Z","iopub.execute_input":"2022-02-26T14:18:22.394523Z","iopub.status.idle":"2022-02-26T14:18:24.219856Z","shell.execute_reply.started":"2022-02-26T14:18:22.394463Z","shell.execute_reply":"2022-02-26T14:18:24.219132Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Among the TOP 100 of solds products:\n- Almost 30% of sold products are trousers\n- 38% is Ladieswear\n- 30% is Lingeries/Tights\n- Over 70% are black colored\n- Almost 40% are related to lower body","metadata":{}},{"cell_type":"markdown","source":"# Q2 - Are there articles that have been sold only once?","metadata":{}},{"cell_type":"markdown","source":"As we observed in the previous analysis, there are items that sold only once. We will investigate about these products.","metadata":{}},{"cell_type":"code","source":"df_sold_qty[\"sold_qty\"].where(lambda x: x==1).dropna() #top 15% products","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:24.221000Z","iopub.execute_input":"2022-02-26T14:18:24.221623Z","iopub.status.idle":"2022-02-26T14:18:24.240599Z","shell.execute_reply.started":"2022-02-26T14:18:24.221590Z","shell.execute_reply":"2022-02-26T14:18:24.239960Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Almost 5000 different items have been sold just once.
\nSince in the \"Transactions\" dataframe there are around 100000 different items, this means that among the transactions, almost 5% of the products have only been sold once.**","metadata":{}},{"cell_type":"markdown","source":"Then we can extract these items from the \"df_sold_qty\" dataframe by taking the last 4491 values. ( There are 4491 items that sold once)","metadata":{}},{"cell_type":"code","source":"worst_sold = df_sold_qty.tail(4491)","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:24.241766Z","iopub.execute_input":"2022-02-26T14:18:24.242010Z","iopub.status.idle":"2022-02-26T14:18:24.245501Z","shell.execute_reply.started":"2022-02-26T14:18:24.241982Z","shell.execute_reply":"2022-02-26T14:18:24.244958Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"And finally join this newly defined dataframe \"worst_sold\" to the articles dataframe df_a to get the articles characterisics.","metadata":{}},{"cell_type":"code","source":"worst_details = sqldf(\"\"\"SELECT *\n FROM worst_sold t\n INNER JOIN df_a a\n on t.article_id = a.article_id\n \"\"\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:24.246654Z","iopub.execute_input":"2022-02-26T14:18:24.247048Z","iopub.status.idle":"2022-02-26T14:18:26.625901Z","shell.execute_reply.started":"2022-02-26T14:18:24.247009Z","shell.execute_reply":"2022-02-26T14:18:26.625048Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(2,2, figsize=(19,14))\nplt.suptitle(\"Characteristic of products sold only once\", size=38, fontweight=\"bold\")\n\nno=100\n\ng = sns.barplot(y=\"product_type_name\", x=\"sold_qty(%)\", data=worst_details.iloc[:no].groupby(\"product_type_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[0,0],palette=\"viridis_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=12, color=\"black\")\nax[0,0].set_ylabel(\"\")\nax[0,0].set_xlabel(\"Sold Quantity (%)\", size=20, fontweight=\"bold\")\nax[0,0].set_title(\"Product Type\", size=25, fontweight=\"bold\")\nax[0,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"index_name\", x=\"sold_qty(%)\", data=worst_details.iloc[:no].groupby(\"index_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[0,1],palette=\"Reds_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15, color=\"black\")\nax[0,1].set_ylabel(\"\")\nax[0,1].set_xlabel(\"Sold Quantity (%)\", size=20, fontweight=\"bold\")\nax[0,1].set_title(\"Index\", size=25, fontweight=\"bold\")\nax[0,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"colour_group_name\", x=\"sold_qty(%)\", data=worst_details.iloc[:no].groupby(\"colour_group_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[1,0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=12, color=\"black\")\nax[1,0].set_ylabel(\"\")\nax[1,0].set_xlabel(\"Sold Quantity (%)\", size=20, fontweight=\"bold\")\nax[1,0].set_title(\"Colour Group\", size=25, fontweight=\"bold\")\nax[1,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"product_group_name\", x=\"sold_qty(%)\", data=worst_details.iloc[:no].groupby(\"product_group_name\")[\"sold_qty\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('sold_qty(%)').reset_index().sort_values(by=\"sold_qty(%)\", ascending=False), \\\n ax=ax[1,1],palette=\"Blues_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15, color=\"black\")\nax[1,1].set_ylabel(\"\")\nax[1,1].set_xlabel(\"Sold Quantity (%)\", size=20, fontweight=\"bold\")\nax[1,1].set_title(\"Product Group\", size=25, fontweight=\"bold\")\nax[1,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\nfig.tight_layout()\n\nplt.show() ","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:26.627470Z","iopub.execute_input":"2022-02-26T14:18:26.628014Z","iopub.status.idle":"2022-02-26T14:18:29.002640Z","shell.execute_reply.started":"2022-02-26T14:18:26.627973Z","shell.execute_reply":"2022-02-26T14:18:29.001664Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Insights:\n- We can see that almost 60% of only sold once items are for children.","metadata":{}},{"cell_type":"markdown","source":"# Q3 - Which are the TOP 100 articles that generated most earnings for the company?","metadata":{}},{"cell_type":"markdown","source":"After analyzing the sold quantites for each product, it can be interesting to analyze the total earnings generated by each product.
\n**The earnings can be calculated by multiplying the price of each product by its total sold quanity**.
\n*NOTE: For privacy reasons, the prices have been transformed/scaled by the creator of the dataset, and so do not represent any known currency.*","metadata":{}},{"cell_type":"markdown","source":"We will now create a new dataframe df_prices which will inlude the earnings generated by each product.","metadata":{}},{"cell_type":"code","source":"df_prices = df_t[[\"price\",\"article_id\"]].groupby(\"article_id\").sum().sort_values(by=\"price\", ascending=False)","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:29.009547Z","iopub.execute_input":"2022-02-26T14:18:29.010118Z","iopub.status.idle":"2022-02-26T14:18:30.713553Z","shell.execute_reply.started":"2022-02-26T14:18:29.010060Z","shell.execute_reply":"2022-02-26T14:18:30.712737Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_prices.rename(columns={\"price\":\"earning\"}, inplace=True)\ndf_prices = df_prices.reset_index()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:30.714690Z","iopub.execute_input":"2022-02-26T14:18:30.714932Z","iopub.status.idle":"2022-02-26T14:18:30.721691Z","shell.execute_reply.started":"2022-02-26T14:18:30.714895Z","shell.execute_reply":"2022-02-26T14:18:30.720769Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_prices.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:30.723236Z","iopub.execute_input":"2022-02-26T14:18:30.723498Z","iopub.status.idle":"2022-02-26T14:18:30.739648Z","shell.execute_reply.started":"2022-02-26T14:18:30.723471Z","shell.execute_reply":"2022-02-26T14:18:30.738794Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**We can see that the most earnings generated by a product is 1631**.
\nHow much is the total earnings?","metadata":{}},{"cell_type":"code","source":"print(\"Number of different sold articles:\",len(df_prices[\"earning\"]))\nprint(\"Total Earnings:\",df_prices[\"earning\"].sum())","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:30.741416Z","iopub.execute_input":"2022-02-26T14:18:30.741741Z","iopub.status.idle":"2022-02-26T14:18:30.754717Z","shell.execute_reply.started":"2022-02-26T14:18:30.741698Z","shell.execute_reply":"2022-02-26T14:18:30.753682Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for i in [10,50,100,200,300,400,1000]:\n print(\"The TOP {} of products that generate most earnings, account for the {:.2f} % of total earnings\".format(i, df_prices[\"earning\"].iloc[:i].sum() / df_prices[\"earning\"].iloc[:].sum() * 100) ) ","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:30.755601Z","iopub.execute_input":"2022-02-26T14:18:30.755837Z","iopub.status.idle":"2022-02-26T14:18:30.770931Z","shell.execute_reply.started":"2022-02-26T14:18:30.755799Z","shell.execute_reply":"2022-02-26T14:18:30.770163Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**The TOP 100 of over 100000 products, generates around 5% of the total earnings. It can be interesting to check these products names and characteristics.**","metadata":{}},{"cell_type":"markdown","source":"So we create a new dataframe top_100_prices, where we include only the TOP 100 articles from the df_prices dataframe.","metadata":{}},{"cell_type":"code","source":"top_100_prices=df_prices.iloc[:100]","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:30.772048Z","iopub.execute_input":"2022-02-26T14:18:30.772725Z","iopub.status.idle":"2022-02-26T14:18:30.777806Z","shell.execute_reply.started":"2022-02-26T14:18:30.772692Z","shell.execute_reply":"2022-02-26T14:18:30.777078Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Then, as seen before, we join this new dataframe to the articles dataframe df_a to get the articles information.","metadata":{}},{"cell_type":"code","source":"top_100_price_details = sqldf(\"\"\"SELECT *\n FROM top_100_prices t\n INNER JOIN df_a a\n on t.article_id = a.article_id\"\"\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:30.778963Z","iopub.execute_input":"2022-02-26T14:18:30.779312Z","iopub.status.idle":"2022-02-26T14:18:33.060009Z","shell.execute_reply.started":"2022-02-26T14:18:30.779284Z","shell.execute_reply":"2022-02-26T14:18:33.059248Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"top_100_price_details.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:33.061288Z","iopub.execute_input":"2022-02-26T14:18:33.062193Z","iopub.status.idle":"2022-02-26T14:18:33.076246Z","shell.execute_reply.started":"2022-02-26T14:18:33.062148Z","shell.execute_reply":"2022-02-26T14:18:33.075399Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(10,11))\nplt.title(\"TOP 50 most profitable products\", size=40, fontweight=\"bold\")\nno=50\ng = sns.barplot(y=\"prod_name\", x=\"earning(%)\", data=top_100_price_details.iloc[:no].groupby(\"prod_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=15)\nplt.xlabel(\"Earnings (%)\", size=25, fontweight=\"bold\")\nplt.ylabel(\"\")\nplt.grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:33.077592Z","iopub.execute_input":"2022-02-26T14:18:33.078118Z","iopub.status.idle":"2022-02-26T14:18:33.990595Z","shell.execute_reply.started":"2022-02-26T14:18:33.078066Z","shell.execute_reply":"2022-02-26T14:18:33.989610Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(2,2, figsize=(13,9))\nplt.suptitle(\"TOP 100 most profitable products characteristics\", fontweight=\"bold\", fontsize=30)\n\nno=100\n\ng = sns.barplot(y=\"product_type_name\", x=\"earning(%)\", data=top_100_price_details.iloc[:no].groupby(\"product_type_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[0,0],palette=\"Blues_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=14, color=\"black\")\nax[0,0].set_ylabel(\"\")\nax[0,0].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[0,0].set_title(\"Product Type\", size=25,fontweight=\"bold\")\nax[0,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\n\ng = sns.barplot(y=\"index_name\", x=\"earning(%)\", data=top_100_price_details.iloc[:no].groupby(\"index_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[0,1],palette=\"viridis\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', padding = 5, fontsize=18, color=\"black\")\nax[0,1].set_ylabel(\"\")\nax[0,1].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[0,1].set_title(\"Index\", size=25,fontweight=\"bold\")\nax[0,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\n\ng = sns.barplot(y=\"colour_group_name\", x=\"earning(%)\", data=top_100_price_details.iloc[:no].groupby(\"colour_group_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[1,0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=18, color=\"black\")\nax[1,0].set_ylabel(\"\")\nax[1,0].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[1,0].set_title(\"Colour Group\", size=25,fontweight=\"bold\")\nax[1,0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\ng = sns.barplot(y=\"product_group_name\", x=\"earning(%)\", data=top_100_price_details.iloc[:no].groupby(\"product_group_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[1,1],palette=\"Reds_r\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', padding=5, fontsize=18, color=\"black\")\nax[1,1].set_ylabel(\"\")\nax[1,1].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[1,1].set_title(\"Product Group\", size=25,fontweight=\"bold\")\nax[1,1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\nfig.tight_layout()\n\nplt.show() ","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:33.992185Z","iopub.execute_input":"2022-02-26T14:18:33.993075Z","iopub.status.idle":"2022-02-26T14:18:35.626490Z","shell.execute_reply.started":"2022-02-26T14:18:33.993031Z","shell.execute_reply":"2022-02-26T14:18:35.625784Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Insights:\n- Over 60% of the TOP 100 products in terms of earnings are generated by selling trousers\n- Around 50% of these products are divided (a H&M teenage collection)\n- 37% of the products are from the Ladieswear line\n- 55% of the products are black\n- 66.2% of the products are related to lower body\n\n**NOTE: It is also important to notice that the TOP 100 most profitable products list do not exactly match the TOP 100 most sold products one, since lots of products that sells a lot in quantity are cheap, and so generate less earnings.**","metadata":{}},{"cell_type":"markdown","source":"# Q4 - Which are articles that generated lower earnings for the company? ","metadata":{}},{"cell_type":"markdown","source":"After checking the most profitable products, it can be interesting to see which are the WORST 100 products in terms of earnings.","metadata":{}},{"cell_type":"code","source":"worst_100_prices=df_prices.iloc[-100:]","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:35.627674Z","iopub.execute_input":"2022-02-26T14:18:35.628517Z","iopub.status.idle":"2022-02-26T14:18:35.632427Z","shell.execute_reply.started":"2022-02-26T14:18:35.628467Z","shell.execute_reply":"2022-02-26T14:18:35.631829Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"worst_100_price_details = sqldf(\"\"\"SELECT *\n FROM worst_100_prices t\n INNER JOIN df_a a\n on t.article_id = a.article_id\"\"\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:35.633524Z","iopub.execute_input":"2022-02-26T14:18:35.634263Z","iopub.status.idle":"2022-02-26T14:18:37.849388Z","shell.execute_reply.started":"2022-02-26T14:18:35.634224Z","shell.execute_reply":"2022-02-26T14:18:37.848620Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"worst_100_price_details.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:37.850669Z","iopub.execute_input":"2022-02-26T14:18:37.851070Z","iopub.status.idle":"2022-02-26T14:18:37.865770Z","shell.execute_reply.started":"2022-02-26T14:18:37.851033Z","shell.execute_reply":"2022-02-26T14:18:37.864458Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(2,1, figsize=(25,15))\nplt.suptitle(\"FLOP 100 Worst profitable products characteristics (1)\", fontsize=40 ,fontweight=\"bold\")\n\nno=100\n\ng = sns.barplot(x=\"product_type_name\", y=\"earning(%)\", data=worst_100_price_details.iloc[:no].groupby(\"product_type_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', fontsize=18)\nax[0].grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nax[0].set_xlabel(\"\")\nax[0].set_ylabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[0].set_xticklabels(g.get_xticklabels(), rotation=80)\nax[0].set_title(\"Product Type\", size=35,fontweight=\"bold\")\n\ng = sns.barplot(x=\"colour_group_name\", y=\"earning(%)\", data=worst_100_price_details.iloc[:no].groupby(\"colour_group_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[1],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', fontsize=18)\nax[1].set_ylabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[1].set_xlabel(\"\")\nax[1].set_xticklabels(g.get_xticklabels(), rotation=80)\nax[1].set_title(\"Colour Group\", size=35,fontweight=\"bold\")\nax[1].grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\n\n\nfig.tight_layout()\n\nplt.show() ","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:37.867320Z","iopub.execute_input":"2022-02-26T14:18:37.867581Z","iopub.status.idle":"2022-02-26T14:18:39.937893Z","shell.execute_reply.started":"2022-02-26T14:18:37.867550Z","shell.execute_reply":"2022-02-26T14:18:39.935914Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Insights: \n- 17.4% of them are t shirts\n- 12.5% are socks\n- There are quite a lot of accessories like hair bands, clips etc..","metadata":{}},{"cell_type":"code","source":"fig, ax = plt.subplots(2,1, figsize=(16,9))\nplt.suptitle(\"FLOP 100 Worst profitable products characteristics (2)\", fontsize=33 ,fontweight=\"bold\")\n\nno=100\n\ng = sns.barplot(y=\"index_name\", x=\"earning(%)\", data=worst_100_price_details.iloc[:no].groupby(\"index_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[0],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', fontsize=18)\nax[0].set_ylabel(\"\")\nax[0].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[0].set_title(\"Index\",size=25,fontweight=\"bold\")\nax[0].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.2)\n \ng = sns.barplot(y=\"product_group_name\", x=\"earning(%)\", data=worst_100_price_details.iloc[:no].groupby(\"product_group_name\")[\"earning\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index().sort_values(by=\"earning(%)\", ascending=False), \\\n ax=ax[1],palette=\"mako\", ci=False)\nfor container in g.containers:\n g.bar_label(container, fmt='%.1f', fontsize=18)\nax[1].set_ylabel(\"\")\nax[1].set_xlabel(\"Earnings (%)\", size=20,fontweight=\"bold\")\nax[1].set_title(\"Product Group\", size=25,fontweight=\"bold\")\nax[1].grid(axis=\"x\",color = 'grey', linestyle = '--', linewidth = 1.5)\n \nplt.tight_layout()\n \nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:18:39.940478Z","iopub.execute_input":"2022-02-26T14:18:39.940921Z","iopub.status.idle":"2022-02-26T14:18:40.825780Z","shell.execute_reply.started":"2022-02-26T14:18:39.940843Z","shell.execute_reply":"2022-02-26T14:18:40.824790Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Insights:\n- 38.7% of these only sonce once products are accessories\n- Around 35% of these products are for children of babies","metadata":{}},{"cell_type":"markdown","source":"# Customer Analayis","metadata":{}},{"cell_type":"markdown","source":"In the following, we will start an analysis on the customers to find interesting insights and understand which customers are responsible for msot purchases.","metadata":{}},{"cell_type":"code","source":"df_t.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:40.827302Z","iopub.execute_input":"2022-02-26T14:18:40.827572Z","iopub.status.idle":"2022-02-26T14:18:40.839260Z","shell.execute_reply.started":"2022-02-26T14:18:40.827538Z","shell.execute_reply":"2022-02-26T14:18:40.837980Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"In order to perform the analysis, we first decide to create a new dataframe that will include, for each row, an unique customer_id, the total purchased quantity by that customer and the ernings generated by the company by the purchases of that customer.","metadata":{}},{"cell_type":"markdown","source":"First, we crate a dataframe which will include the unique customer ids and the earnings generated by theirs purchases.","metadata":{}},{"cell_type":"code","source":"df_cust_prices = df_t[[\"customer_id\", \"price\"]].groupby(\"customer_id\").sum()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:40.840783Z","iopub.execute_input":"2022-02-26T14:18:40.841857Z","iopub.status.idle":"2022-02-26T14:18:55.350785Z","shell.execute_reply.started":"2022-02-26T14:18:40.841738Z","shell.execute_reply":"2022-02-26T14:18:55.349843Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_cust_prices.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:55.352177Z","iopub.execute_input":"2022-02-26T14:18:55.353054Z","iopub.status.idle":"2022-02-26T14:18:55.362721Z","shell.execute_reply.started":"2022-02-26T14:18:55.353009Z","shell.execute_reply":"2022-02-26T14:18:55.361422Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Second, we create a dataframe that will include the unique customer ids and their total purchased quantity of products.","metadata":{}},{"cell_type":"code","source":"df_cust_qty = df_t[[\"customer_id\", \"article_id\"]].groupby(\"customer_id\").count()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:18:55.364118Z","iopub.execute_input":"2022-02-26T14:18:55.364480Z","iopub.status.idle":"2022-02-26T14:19:09.407653Z","shell.execute_reply.started":"2022-02-26T14:18:55.364446Z","shell.execute_reply":"2022-02-26T14:19:09.406795Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_cust_qty.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:09.408835Z","iopub.execute_input":"2022-02-26T14:19:09.409072Z","iopub.status.idle":"2022-02-26T14:19:09.419049Z","shell.execute_reply.started":"2022-02-26T14:19:09.409045Z","shell.execute_reply":"2022-02-26T14:19:09.418137Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Then, we join these two dataframe to a new one \"cust_qty_price\", which will include the unique customer ids, their purchased quantity and the earnings generated by the company by their purchases.","metadata":{}},{"cell_type":"code","source":"cust_qty_price = pd.merge(df_cust_prices, df_cust_qty, on='customer_id', how='inner')","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:09.420500Z","iopub.execute_input":"2022-02-26T14:19:09.420803Z","iopub.status.idle":"2022-02-26T14:19:11.056732Z","shell.execute_reply.started":"2022-02-26T14:19:09.420771Z","shell.execute_reply":"2022-02-26T14:19:11.056029Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_qty_price.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:11.057834Z","iopub.execute_input":"2022-02-26T14:19:11.058531Z","iopub.status.idle":"2022-02-26T14:19:11.067809Z","shell.execute_reply.started":"2022-02-26T14:19:11.058497Z","shell.execute_reply":"2022-02-26T14:19:11.067083Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Finally we can join this new dataframe to the Customer dataframe df_c, so that we can add some informations about the customer on the newly defined cust_qty_price dataframe.","metadata":{}},{"cell_type":"code","source":"df_c.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:11.068887Z","iopub.execute_input":"2022-02-26T14:19:11.069212Z","iopub.status.idle":"2022-02-26T14:19:11.093540Z","shell.execute_reply.started":"2022-02-26T14:19:11.069183Z","shell.execute_reply":"2022-02-26T14:19:11.092926Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_details = pd.merge(cust_qty_price, df_c.drop(\"postal_code\", axis=1), on='customer_id', how='inner')","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:11.095034Z","iopub.execute_input":"2022-02-26T14:19:11.095530Z","iopub.status.idle":"2022-02-26T14:19:13.321978Z","shell.execute_reply.started":"2022-02-26T14:19:11.095496Z","shell.execute_reply":"2022-02-26T14:19:13.321035Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cust_details.head()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:13.323269Z","iopub.execute_input":"2022-02-26T14:19:13.323601Z","iopub.status.idle":"2022-02-26T14:19:13.339905Z","shell.execute_reply.started":"2022-02-26T14:19:13.323558Z","shell.execute_reply":"2022-02-26T14:19:13.339031Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f\"In total there are {len(cust_details)} different customers\")","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:13.341408Z","iopub.execute_input":"2022-02-26T14:19:13.341931Z","iopub.status.idle":"2022-02-26T14:19:13.355564Z","shell.execute_reply.started":"2022-02-26T14:19:13.341865Z","shell.execute_reply":"2022-02-26T14:19:13.354447Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Purchased Quantity by Customer Analysis","metadata":{"execution":{"iopub.status.busy":"2022-02-24T08:46:46.6228Z","iopub.execute_input":"2022-02-24T08:46:46.623519Z","iopub.status.idle":"2022-02-24T08:46:46.631323Z","shell.execute_reply.started":"2022-02-24T08:46:46.623472Z","shell.execute_reply":"2022-02-24T08:46:46.629881Z"}}},{"cell_type":"markdown","source":"Now we will analyze the purchased quantity by the customers.","metadata":{}},{"cell_type":"code","source":"cust_details.article_id.describe()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:13.356774Z","iopub.execute_input":"2022-02-26T14:19:13.357398Z","iopub.status.idle":"2022-02-26T14:19:13.418107Z","shell.execute_reply.started":"2022-02-26T14:19:13.357352Z","shell.execute_reply":"2022-02-26T14:19:13.417248Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"By calling the \"describe\" method on the \"article_id\" column, we can observe that:\n- The minimum purchased quantity by a single customer is 1\n- 25% of customers Purchased 3 or less items\n- 50% of customers Purchased 9 or less items\n- 75% of customers Purchased 27 or less items\n- The maximum purchased quantity by a single customer is 1895 products","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(10,4))\nplt.title(\"Distribution of purchased quantity by customer\", fontweight=\"bold\", size=20)\nsns.kdeplot(cust_details[\"article_id\"])\nplt.xlabel(\"purchased quantity\",fontweight=\"bold\", size=20)\nplt.ylabel(\"Count\",fontweight=\"bold\", size=20)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:13.419475Z","iopub.execute_input":"2022-02-26T14:19:13.419953Z","iopub.status.idle":"2022-02-26T14:19:18.221390Z","shell.execute_reply.started":"2022-02-26T14:19:13.419919Z","shell.execute_reply":"2022-02-26T14:19:18.220545Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Indeed, the distribution of this variable is highly skewed.","metadata":{}},{"cell_type":"markdown","source":"Next, we will analyze the age and other provided features of the customer to better find insights on the customers and their purchase behaviour.","metadata":{}},{"cell_type":"markdown","source":"# Purchase Behaviors according to Age","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(10,5))\nplt.title(\"Customers age distribution\", fontweight=\"bold\", size=30)\nplt.hist(cust_details[\"age\"], bins=70, edgecolor=\"black\", color=\"#1ABC9C\")\nplt.xlabel(\"Age\",fontweight=\"bold\", size=20)\nplt.ylabel(\"Count\",fontweight=\"bold\", size=20)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:18.222727Z","iopub.execute_input":"2022-02-26T14:19:18.223538Z","iopub.status.idle":"2022-02-26T14:19:18.681987Z","shell.execute_reply.started":"2022-02-26T14:19:18.223490Z","shell.execute_reply":"2022-02-26T14:19:18.681163Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**The distribution of the age feature is bivariate. In order to create more effective plots, we will create a categorical column for age which divides the ages in age groups.**","metadata":{}},{"cell_type":"markdown","source":"# Q5 - Which age group purchase more articles?","metadata":{}},{"cell_type":"code","source":"cust_details['age_groups'] = pd.cut(cust_details['age'], bins=[16, 20, 30, 40,50, 60, 70, float('Inf')], labels=['16-20', '20-30','30-40','40-50','50-60','60-70' , '70+'])","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:18.683398Z","iopub.execute_input":"2022-02-26T14:19:18.683707Z","iopub.status.idle":"2022-02-26T14:19:18.737590Z","shell.execute_reply.started":"2022-02-26T14:19:18.683667Z","shell.execute_reply":"2022-02-26T14:19:18.736773Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(8,5))\nplt.title(\"Purchased quantity by age group\\n\", fontweight=\"bold\", size=28)\ng = sns.barplot(x=\"age_groups\", y=\"Purchased Quantity(%)\", data=cust_details.groupby(\"age_groups\")[\"article_id\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('Purchased Quantity(%)').reset_index(), palette=\"icefire\", edgecolor=\"black\")\nplt.xlabel(\"Age Group\",fontweight=\"bold\", size=22)\nplt.ylabel(\"Purchased Quantity (%)\",fontweight=\"bold\", size=19)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=18, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:18.739097Z","iopub.execute_input":"2022-02-26T14:19:18.739895Z","iopub.status.idle":"2022-02-26T14:19:19.118839Z","shell.execute_reply.started":"2022-02-26T14:19:18.739828Z","shell.execute_reply":"2022-02-26T14:19:19.117965Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Insights:\n- Customers in the range 20-30 are responsible for more than 42% of the total purchased products.\n- Customers in the range 16-20. 60-70 and 70+ are responsible for the 8% of the total purchased products\n- Customers in the range 30-40, 40-50 and 50-60 are responsible for 16% of purchased quantity each.","metadata":{}},{"cell_type":"markdown","source":"After analyzing the purchases quantity, it could be interesting to analyze the earnings provided to the company by each customer.","metadata":{"execution":{"iopub.status.busy":"2022-02-24T09:09:34.101437Z","iopub.execute_input":"2022-02-24T09:09:34.101783Z","iopub.status.idle":"2022-02-24T09:09:34.108475Z","shell.execute_reply.started":"2022-02-24T09:09:34.101748Z","shell.execute_reply":"2022-02-24T09:09:34.107457Z"}}},{"cell_type":"markdown","source":"# Q6 - Which age group generates more earnings for the company?","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(8,5))\nplt.title(\"Company Earnings by age group\\n\", fontweight=\"bold\", size=28)\ng = sns.barplot(x=\"age_groups\", y=\"earning(%)\", data=cust_details.groupby(\"age_groups\")[\"price\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('earning(%)').reset_index(), palette=\"icefire\",edgecolor=\"black\")\nplt.xlabel(\"Age Group\",fontweight=\"bold\", size=22)\nplt.ylabel(\"Earnings (%)\",fontweight=\"bold\", size=25)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=18, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:19.120402Z","iopub.execute_input":"2022-02-26T14:19:19.120917Z","iopub.status.idle":"2022-02-26T14:19:19.486290Z","shell.execute_reply.started":"2022-02-26T14:19:19.120854Z","shell.execute_reply":"2022-02-26T14:19:19.485454Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n**Indeed a very similar situation to the purchases quantity can be found in the earnings analysis, since customers who buys more, on average leads to higher earnings for the company.
\nThe age group 20-30 is by far responsible for the highest earnings for the company (41.9% of total earnings).**","metadata":{}},{"cell_type":"markdown","source":"# Q7 - Do active customers on the fashion news purchase more articles?","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(9,5))\nplt.title(\"Purchased quantity by Fashion News Frequency\\n\", fontweight=\"bold\", size=20)\ng = sns.barplot(x=\"fashion_news_frequency\", y=\"Purchased Quantity(%)\", data=cust_details.groupby(\"fashion_news_frequency\")[\"article_id\"].sum() \\\n .transform(lambda x: (x / x.sum() * 100)).rename('Purchased Quantity(%)').reset_index(), palette=\"Spectral\", edgecolor=\"black\")\nplt.xlabel(\"Fashion News Frequency\",fontweight=\"bold\", size=22)\nplt.ylabel(\"Purchased Quantity (%)\",fontweight=\"bold\", size=25)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.3f', fontsize=18, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:19.487795Z","iopub.execute_input":"2022-02-26T14:19:19.488293Z","iopub.status.idle":"2022-02-26T14:19:19.969569Z","shell.execute_reply.started":"2022-02-26T14:19:19.488249Z","shell.execute_reply":"2022-02-26T14:19:19.968622Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Active customers on the fashion news are responsible for 43% of the total purchases, while the remaining 57% of purchased quantity comes from customer not registed in the fashion news.**
\nThe other 2 categories \"Monthly\" and \"None\" can be ignored and won't be considered for the further analysis.","metadata":{}},{"cell_type":"markdown","source":"So then it could be interesting to check the fashion news frequency by age group, to find more useful insights,","metadata":{}},{"cell_type":"code","source":"x, y = 'age_groups', 'fashion_news_frequency'\ndf_age_news = cust_details.groupby(x)[y].value_counts(normalize=True)\ndf_age_news = df_age_news.mul(100)\ndf_age_news = df_age_news.rename('percent(%)').reset_index()\ndf_age_news = df_age_news[df_age_news[\"fashion_news_frequency\"].isin([\"Regularly\",\"NONE\"])]","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:19.971064Z","iopub.execute_input":"2022-02-26T14:19:19.971412Z","iopub.status.idle":"2022-02-26T14:19:20.345379Z","shell.execute_reply.started":"2022-02-26T14:19:19.971368Z","shell.execute_reply":"2022-02-26T14:19:20.344469Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"palette1 = {\"Regularly\":'#46C646', \"NONE\":'#FF0000'}\n\nplt.figure(figsize=(13,6))\nplt.title(\"Fashion News Frequency by age group\\n\",fontweight=\"bold\", size=33)\ng=sns.barplot(x=\"age_groups\", y=\"percent(%)\",data=df_age_news, hue=\"fashion_news_frequency\", palette=palette1)\nplt.xlabel(\"Age group\",fontweight=\"bold\", size=22)\nplt.ylabel(\"Percentage (%)\",fontweight=\"bold\", size=25)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.1f', fontsize=16, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.legend(title='News\\nFrequency',bbox_to_anchor=(1.0, 1.0), ncol=1, fancybox=True, shadow=True, fontsize=17,title_fontsize=22)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:20.346658Z","iopub.execute_input":"2022-02-26T14:19:20.346963Z","iopub.status.idle":"2022-02-26T14:19:20.853405Z","shell.execute_reply.started":"2022-02-26T14:19:20.346931Z","shell.execute_reply":"2022-02-26T14:19:20.852506Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"We can see that **customers in the range 20-30 and 30-40 have the lowest percentage of fashion news frequency, while being the groups which buy the most**.
\nMoreover, **the frequency of customer that regulary check fashion news starts increasing from the range 40-50, with a peak value of 43.7% of regular/active users for customers in the range 70+ years old**. This means that **checking fashion news seems to be more effective for older customers, who still represent a small percentage of total sold products, while younger customers do not need to check the news to buy new products**.
\n**It could be effective for the company to invite younger customers (range 20-40) to check the news more frequently in order to increase the sold items.**","metadata":{}},{"cell_type":"markdown","source":"# Q8 - Does the club member status influence the purchased quantity?","metadata":{}},{"cell_type":"code","source":"cust_details[\"club_member_status\"].value_counts(normalize=True)","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:20.854690Z","iopub.execute_input":"2022-02-26T14:19:20.855226Z","iopub.status.idle":"2022-02-26T14:19:21.076477Z","shell.execute_reply.started":"2022-02-26T14:19:20.855167Z","shell.execute_reply":"2022-02-26T14:19:21.075439Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"We can see that:\n- More than 93% of the customers belong to the ACTIVE category\n- 6.8% of the customers belong to the PRE-CREATE cateory\n- 0.3% of the customers belong to the LEFT CLUB category","metadata":{}},{"cell_type":"markdown","source":"**This shows a very high imbalance among the classes: if we consider the sum of purchased products per each category, this will likely show that the most part of Purchased products belongs to the ACTIVE members.**","metadata":{}},{"cell_type":"code","source":"cust_details.groupby(\"club_member_status\")[\"article_id\"].sum()","metadata":{"execution":{"iopub.status.busy":"2022-02-26T14:19:21.077748Z","iopub.execute_input":"2022-02-26T14:19:21.078002Z","iopub.status.idle":"2022-02-26T14:19:21.282776Z","shell.execute_reply.started":"2022-02-26T14:19:21.077972Z","shell.execute_reply":"2022-02-26T14:19:21.281661Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Indeed, more customers in a group leads to higher purchases. For this reason, it is more wise to consider a mean Purchased quantity instead of a sum:**","metadata":{}},{"cell_type":"code","source":"print(\"The average quantity of purchased products by the customers is {:.0f} products \".format(cust_details[\"article_id\"].mean()))","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:21.284227Z","iopub.execute_input":"2022-02-26T14:19:21.284511Z","iopub.status.idle":"2022-02-26T14:19:21.292236Z","shell.execute_reply.started":"2022-02-26T14:19:21.284471Z","shell.execute_reply":"2022-02-26T14:19:21.290955Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"The average quantity of purchased products by the ACTIVE customers is {:.0f} products \".format(cust_details.groupby(\"club_member_status\")[\"article_id\"].mean()[\"ACTIVE\"]))\nprint(\"The average quantity of purchased products by the LEFT-CLUB customers is {:.0f} products \".format(cust_details.groupby(\"club_member_status\")[\"article_id\"].mean()[\"LEFT CLUB\"]))\nprint(\"The average quantity of purchased products by the PRE-CREATE customers is {:.0f} products \".format(cust_details.groupby(\"club_member_status\")[\"article_id\"].mean()[\"PRE-CREATE\"]))","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:21.293549Z","iopub.execute_input":"2022-02-26T14:19:21.293814Z","iopub.status.idle":"2022-02-26T14:19:21.887827Z","shell.execute_reply.started":"2022-02-26T14:19:21.293773Z","shell.execute_reply":"2022-02-26T14:19:21.886842Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"By considering the mean, we can see a very different situation, which will be shown as percentages in the following plot:","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(9,5))\nplt.title(\"Average Purchased Quantity by Club Member Status\\n\", fontweight=\"bold\", size=22)\ng = sns.barplot(x=\"club_member_status\", y=\"article_id\", data=cust_details.groupby(\"club_member_status\")[\"article_id\"].mean().astype(int).reset_index(), palette=\"viridis\", edgecolor=\"black\")\nplt.axhline(y = cust_details[\"article_id\"].mean(), color = 'r', linestyle = '--')\nplt.text(0.76, 23.7, 'Mean Purchased Quantity: {:.0f}'.format(cust_details[\"article_id\"].mean()), size=16, color=\"red\",fontweight=\"bold\")\nplt.xlabel(\"Club Member Status\",fontweight=\"bold\", size=20)\nplt.ylabel(\"Average Purchased Quantity\",fontweight=\"bold\", size=16)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.0f', fontsize=23, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:21.889094Z","iopub.execute_input":"2022-02-26T14:19:21.889352Z","iopub.status.idle":"2022-02-26T14:19:22.376078Z","shell.execute_reply.started":"2022-02-26T14:19:21.889323Z","shell.execute_reply":"2022-02-26T14:19:22.375353Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**This plots shows that the average purchased quantity differs a lot among the categories**.
\nIn particular, **customers belonging to the ACTIVE clubs, purchase more products than other categories, while those in the \"pre-create\" category purchaes on average less than a third of third of active customers**.","metadata":{}},{"cell_type":"markdown","source":"Finally, since the distribution of the purchased quantity is heavily right skewed, it could be interesting to check out also the median purhcased quantity.","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(9,5))\nplt.title(\"Median Purchased Quantity by Club Member Status\\n\", fontweight=\"bold\", size=22)\ng = sns.barplot(x=\"club_member_status\", y=\"article_id\", data=cust_details.groupby(\"club_member_status\")[\"article_id\"].median().reset_index(), palette=\"viridis\", edgecolor=\"black\")\nplt.axhline(y = cust_details[\"article_id\"].median(), color = 'r', linestyle = '--')\nplt.text(0.76, 9.3, 'Median Purchased Quantity: {:.2f}'.format(cust_details[\"article_id\"].median()), size=16, color=\"red\",fontweight=\"bold\")\nplt.xlabel(\"Club Member Status\",fontweight=\"bold\", size=20)\nplt.ylabel(\"Median Purchaed Quantity\",fontweight=\"bold\", size=16)\nfor container in g.containers:\n g.bar_label(container, padding = 5, fmt='%.0f', fontsize=23, color=\"black\")\nplt.grid(axis=\"y\",color = 'grey', linestyle = '--', linewidth = 1.5)\nplt.show()","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-02-26T14:19:22.377254Z","iopub.execute_input":"2022-02-26T14:19:22.377606Z","iopub.status.idle":"2022-02-26T14:19:22.911743Z","shell.execute_reply.started":"2022-02-26T14:19:22.377575Z","shell.execute_reply":"2022-02-26T14:19:22.910771Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Indeed, even if the Median is quite different for the Mean due to high skeweness of the data, a very similar situation situation to the mean purchases quantity can be observed, where ACTIVE customers buys more product on average.","metadata":{}},{"cell_type":"markdown","source":"**This Notebook is still a W.I.P., I will update it with predictions or more data analysis if I have new ideas !!**","metadata":{}},{"cell_type":"markdown","source":"**Thank your for checking out my notebook! Let me know if you have comments or if you want me to check out your work! :)**","metadata":{}}]}