Project 2 .ipynb

{"metadata":{"orig_nbformat":4,"colab":{"collapsed_sections":["q491HYdbIq8T","jkdo7SE6Mw3N","y3OJYgdqN38Y","Zfz0fzTpA-3j"],"name":"Assignment2.ipynb","provenance":[],"version":"0.3.2"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.7.8","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# Project 2","metadata":{"colab_type":"text","id":"ZiWHjXgvEChM"}},{"cell_type":"markdown","source":"# Preprocessing and Data Quality \n","metadata":{"colab_type":"text","id":"3pftdbPtBam4"}},{"cell_type":"markdown","source":"### Question 1 (Order cancellations)\nInvoices with a InvoiceNo starting with the letter ‘c’ are order cancellations. Would you recommend keeping the order cancellation in your data set? Also provide a reason for your recommandation. ","metadata":{"colab_type":"text","id":"fLUecg7SFaoG"}},{"cell_type":"markdown","source":"Your explanation:\n- As mentioned in the description of the dataset, some InvoiceNo start with the letter \"c\" = cancelled. By looking at the list of cancelled invoices, there are no positive value for quantity where the invoice is prefixed with 'c'. There are 3836 unique orders that are cancelled out of total 25900 (Unique InvoiceNo).It is removing only 1.7% of total data. I recommend removing it since there are no specific analysis in the further questions based on cancellation data.","metadata":{"colab_type":"text","id":"xkSmAfDFFycc"}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport csv\nfrom matplotlib import pyplot as plt\nimport pandas as pd\nimport seaborn as sns\nimport dateutil\nimport warnings\n\n%matplotlib inline\nwarnings.filterwarnings(\"ignore\")\n\n\ndataframe = pd.read_excel(\"OnlineRetail.xlsx\")\n\n#getting cancelled transactions with unique InvoiceNo.\ncancelled_orders = dataframe[dataframe['InvoiceNo'].astype(str).str.contains('C')]\ntotal_cancelled_orders =cancelled_orders['InvoiceNo'].nunique()\nprint(total_cancelled_orders)","metadata":{},"execution_count":1,"outputs":[{"name":"stdout","output_type":"stream","text":"3836\n"}]},{"cell_type":"code","source":"cancelled_orders.shape","metadata":{},"execution_count":2,"outputs":[{"execution_count":2,"output_type":"execute_result","data":{"text/plain":"(9288, 8)"},"metadata":{}}]},{"cell_type":"code","source":"#Number of total orders with unique InvoiceNo.\ntotal_orders = dataframe['InvoiceNo'].nunique()\nprint(total_orders)","metadata":{},"execution_count":3,"outputs":[{"name":"stdout","output_type":"stream","text":"25900\n"}]},{"cell_type":"code","source":"#cancelled orders with quantity greater than 0\ncancelled_orders[cancelled_orders['Quantity']>0]","metadata":{},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>InvoiceNo</th>\n      <th>StockCode</th>\n      <th>Description</th>\n      <th>Quantity</th>\n      <th>InvoiceDate</th>\n      <th>UnitPrice</th>\n      <th>CustomerID</th>\n      <th>Country</th>\n    </tr>\n  </thead>\n  <tbody>\n  </tbody>\n</table>\n</div>","text/plain":"Empty DataFrame\nColumns: [InvoiceNo, StockCode, Description, Quantity, InvoiceDate, UnitPrice, CustomerID, Country]\nIndex: []"},"metadata":{}}]},{"cell_type":"code","source":"# Removing Cancelled Orders\ndataframe_removed_cancelled_orders=dataframe[~dataframe.InvoiceNo.str.contains(\"C\", na=False)]\nprint('Initial dataframe shape:',dataframe.shape)\nprint('After removal dataframe shape:',dataframe_removed_cancelled_orders.shape)","metadata":{},"execution_count":5,"outputs":[{"name":"stdout","output_type":"stream","text":"Initial dataframe shape: (541909, 8)\nAfter removal dataframe shape: (532621, 8)\n"}]},{"cell_type":"markdown","source":"### Question 2 (Empty values)\nThe attributes Description and CustomerID contain empty values. The Country attribute contains an “unspecified” value. For each of the three attributes reason how you would handle these values and why. ","metadata":{"colab_type":"text","id":"OugGYUj3HCcD"}},{"cell_type":"markdown","source":"Your explanation:\n- Description: Description missing values is not handled since Stockcode can be used to get the description and description is not used in Clustering and Frequent Item sets. \n- CustomerID: Missing values in CustomerID is not handled because the Clustering and Frequent Item sets are not based on the product bought. Removing all the missing values of CustomerID will also lead to a removal of 25% of data rows.\n- Country: Replacing the unspecified with mode is suitable as 486732/532621 availible dataset has United Kingdom. So replacing 446 unspecified rows with United Kingdom will not change the dataset significantly ","metadata":{"colab_type":"text","id":"4VQ4ArG8IYhE"}},{"cell_type":"code","source":"dataframe_removed_cancelled_orders.info()\n\n#preprocessing missing Country\n\nunspecified =  dataframe_removed_cancelled_orders['Country']==\"Unspecified\"\nunspecified_dataframe = dataframe_removed_cancelled_orders[unspecified]\nprint(unspecified_dataframe.shape) #number of entries with unspecified\n\ncountry_mode = dataframe_removed_cancelled_orders['Country'].mode()[0]\ndataframe_removed_cancelled_orders['Country'] = dataframe_removed_cancelled_orders['Country'].replace(\"Unspecified\",country_mode)","metadata":{"colab":{},"colab_type":"code","id":"k8c855xlILgG"},"execution_count":6,"outputs":[{"name":"stdout","output_type":"stream","text":"<class 'pandas.core.frame.DataFrame'>\nInt64Index: 532621 entries, 0 to 541908\nData columns (total 8 columns):\nInvoiceNo      532621 non-null object\nStockCode      532621 non-null object\nDescription    531167 non-null object\nQuantity       532621 non-null int64\nInvoiceDate    532621 non-null datetime64[ns]\nUnitPrice      532621 non-null float64\nCustomerID     397924 non-null float64\nCountry        532621 non-null object\ndtypes: datetime64[ns](1), float64(2), int64(1), object(4)\nmemory usage: 36.6+ MB\n(446, 8)\n"}]},{"cell_type":"markdown","source":"### Question 3 (Outliers/Noise)\nExplore into the attributes Quantity and UnitPrice by plotting each attribute visually. Do these attributes contain noise and/or outliers? If so, reason how you would handle them and modify your data set accordingly.\n","metadata":{"colab_type":"text","id":"q491HYdbIq8T"}},{"cell_type":"code","source":"#Boxplotting the Quantity and UnitPrice\nred_square = dict(markerfacecolor='r', marker='s')\nfig5, ax5 = plt.subplots()\nax5.set_title('BoxPlot of Quantity and UnitPrice')\ndataframe_removed_cancelled_orders.boxplot(column=[\"Quantity\",\"UnitPrice\"],flierprops=red_square) \nplt.show()","metadata":{},"execution_count":7,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182c59e67b8>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"# #Method to Remove Outliers in quantity and unitprice\n\ndatafram_outliers_removed=dataframe_removed_cancelled_orders\n\n\ndef remove_outlier(df_in, col_name):\n    q1 = df_in[col_name].quantile(0.25)\n    q3 = df_in[col_name].quantile(0.75)\n    iqr = q3-q1 #Interquartile range\n    fence_low  = q1-1.5*iqr\n    fence_high = q3+1.5*iqr\n    df_out = df_in.loc[(df_in[col_name] > fence_low) & (df_in[col_name] < fence_high)]\n    return df_out\n\ndata_filtered_quantity = remove_outlier(datafram_outliers_removed,'Quantity')\n\ndataframe_outliers_removed = remove_outlier(data_filtered_quantity,'UnitPrice')\n\ndataframe_outliers_removed.shape","metadata":{"colab":{},"colab_type":"code","id":"-bOdS-GsJPn3"},"execution_count":8,"outputs":[{"execution_count":8,"output_type":"execute_result","data":{"text/plain":"(438509, 8)"},"metadata":{}}]},{"cell_type":"code","source":"fig5, ax5 = plt.subplots()\ngreen_diamond = dict(markerfacecolor='g', marker='D')\nax5.set_title('After Removing Outliers Box plot of Quantity')\ndataframe_outliers_removed.boxplot(column=[\"Quantity\"],flierprops=green_diamond)\nplt.show()","metadata":{},"execution_count":9,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e891b978>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"fig5, ax5 = plt.subplots()\ngreen_diamond = dict(markerfacecolor='g', marker='D')\nax5.set_title('After Removing Outliers Box plot of Hours-per-week')\ndataframe_outliers_removed.boxplot(column=[\"UnitPrice\"],flierprops=green_diamond)\nplt.show()","metadata":{},"execution_count":10,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e42aea20>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"Removal of rows with negative quantity","metadata":{}},{"cell_type":"markdown","source":"Note: Found while exporing data, the above dataset has quantity in negative. Together with that both CustomerID, UnitPrice are unavailible. Description is availible for partial dataset. So it is clear that the dataset does not represent purchases. So, I plan to remove them also.","metadata":{}},{"cell_type":"code","source":"# dataset with negative quantity\nprint(dataframe_outliers_removed[dataframe_outliers_removed[\"Quantity\"] < 0])","metadata":{},"execution_count":11,"outputs":[{"name":"stdout","output_type":"stream","text":"       InvoiceNo StockCode Description  Quantity         InvoiceDate  \\\n2406      536589     21777         NaN       -10 2010-12-01 16:50:00   \n7190      536998     85067         NaN        -6 2010-12-03 15:30:00   \n7193      537001     21653         NaN        -6 2010-12-03 15:33:00   \n7195      537003     85126         NaN        -2 2010-12-03 15:33:00   \n7203      537011     22139         NaN        -5 2010-12-03 15:38:00   \n7204      537012     21535         NaN       -12 2010-12-03 15:39:00   \n7210      537018    84562A         NaN        -3 2010-12-03 15:42:00   \n7293      537029     37476         NaN        -6 2010-12-03 16:37:00   \n14062     537451     84881         NaN       -11 2010-12-07 09:27:00   \n17158     537687     84840         NaN        -9 2010-12-08 10:26:00   \n20393     538042     21763         NaN        -4 2010-12-09 13:10:00   \n20394     538043     22833         NaN        -1 2010-12-09 13:10:00   \n31226     538906     21895         NaN        -5 2010-12-15 10:31:00   \n42498     539994     22830         NaN        -1 2011-01-04 10:09:00   \n42499     539995     85064         NaN       -12 2011-01-04 10:12:00   \n42500     539996     21784         NaN       -10 2011-01-04 10:12:00   \n42503     539999     20892         NaN        -1 2011-01-04 10:16:00   \n42504     540000     84499         NaN        -9 2011-01-04 10:17:00   \n51048     540637     21275         NaN        -5 2011-01-10 12:13:00   \n51672     540650    85124C         NaN        -8 2011-01-10 14:59:00   \n51762     540678     21667         NaN        -1 2011-01-10 16:21:00   \n51763     540679     21667         NaN        -1 2011-01-10 16:21:00   \n69215     541967     84823         NaN        -3 2011-01-24 13:35:00   \n73850     542392     84227         NaN        -2 2011-01-27 14:56:00   \n73935     542396     22146         NaN        -4 2011-01-27 15:12:00   \n74283     542414    84967A         NaN        -2 2011-01-27 17:15:00   \n74284     542415    84967B         NaN        -1 2011-01-27 17:15:00   \n74287     542417    84966B         NaN       -11 2011-01-27 17:38:00   \n74825     542529  DCGS0055         NaN        -1 2011-01-28 13:08:00   \n74838     542531  DCGS0072         NaN        -1 2011-01-28 13:08:00   \n...          ...       ...         ...       ...                 ...   \n422821    573120     72816         NaN        -1 2011-10-27 16:21:00   \n432117    573815     20902       check        -3 2011-11-01 11:31:00   \n432118    573816     20898       check       -10 2011-11-01 11:32:00   \n436998    574257     72818         NaN        -3 2011-11-03 13:46:00   \n440212    574520     21181         NaN        -6 2011-11-04 13:44:00   \n453963    575503     35933         NaN        -2 2011-11-10 10:19:00   \n455405    575613     23118       check        -6 2011-11-10 12:47:00   \n455406    575614     22121       check       -10 2011-11-10 12:48:00   \n455512    575625     35972       check        -3 2011-11-10 13:20:00   \n456829    575720     21817       check        -8 2011-11-10 18:09:00   \n460357    575907     22268       check       -10 2011-11-11 15:52:00   \n460398    575912     22943       check        -7 2011-11-11 16:20:00   \n464972    576193     22827         NaN        -8 2011-11-14 11:57:00   \n468300    576415    17012B         NaN        -2 2011-11-15 11:22:00   \n468302    576424    85036C       check        -4 2011-11-15 11:24:00   \n469110    576580     22034         NaN        -5 2011-11-15 14:15:00   \n472222    576674     21823     damaged        -6 2011-11-16 12:00:00   \n473440    576835     21618         NaN        -2 2011-11-16 15:10:00   \n478932    577119     22778       check       -10 2011-11-17 18:08:00   \n478933    577120     20752       check       -10 2011-11-17 18:09:00   \n492198    578095     23397     damaged        -9 2011-11-22 17:25:00   \n497569    578426     23131         NaN       -12 2011-11-24 12:03:00   \n500145    578733     21643     damaged        -9 2011-11-25 11:32:00   \n514287    579691    47503A         NaN        -5 2011-11-30 14:05:00   \n514647    579732     35648       check       -10 2011-11-30 14:24:00   \n517719    580087    84341B         NaN        -2 2011-12-01 14:27:00   \n522162    580381     21758         NaN        -9 2011-12-02 17:58:00   \n524472    580561     22043      dotcom        -9 2011-12-05 10:25:00   \n535322    581199     84581         NaN        -2 2011-12-07 18:26:00   \n535331    581208    72801C       check       -10 2011-12-07 18:35:00   \n\n        UnitPrice  CustomerID         Country  \n2406          0.0         NaN  United Kingdom  \n7190          0.0         NaN  United Kingdom  \n7193          0.0         NaN  United Kingdom  \n7195          0.0         NaN  United Kingdom  \n7203          0.0         NaN  United Kingdom  \n7204          0.0         NaN  United Kingdom  \n7210          0.0         NaN  United Kingdom  \n7293          0.0         NaN  United Kingdom  \n14062         0.0         NaN  United Kingdom  \n17158         0.0         NaN  United Kingdom  \n20393         0.0         NaN  United Kingdom  \n20394         0.0         NaN  United Kingdom  \n31226         0.0         NaN  United Kingdom  \n42498         0.0         NaN  United Kingdom  \n42499         0.0         NaN  United Kingdom  \n42500         0.0         NaN  United Kingdom  \n42503         0.0         NaN  United Kingdom  \n42504         0.0         NaN  United Kingdom  \n51048         0.0         NaN  United Kingdom  \n51672         0.0         NaN  United Kingdom  \n51762         0.0         NaN  United Kingdom  \n51763         0.0         NaN  United Kingdom  \n69215         0.0         NaN  United Kingdom  \n73850         0.0         NaN  United Kingdom  \n73935         0.0         NaN  United Kingdom  \n74283         0.0         NaN  United Kingdom  \n74284         0.0         NaN  United Kingdom  \n74287         0.0         NaN  United Kingdom  \n74825         0.0         NaN  United Kingdom  \n74838         0.0         NaN  United Kingdom  \n...           ...         ...             ...  \n422821        0.0         NaN  United Kingdom  \n432117        0.0         NaN  United Kingdom  \n432118        0.0         NaN  United Kingdom  \n436998        0.0         NaN  United Kingdom  \n440212        0.0         NaN  United Kingdom  \n453963        0.0         NaN  United Kingdom  \n455405        0.0         NaN  United Kingdom  \n455406        0.0         NaN  United Kingdom  \n455512        0.0         NaN  United Kingdom  \n456829        0.0         NaN  United Kingdom  \n460357        0.0         NaN  United Kingdom  \n460398        0.0         NaN  United Kingdom  \n464972        0.0         NaN  United Kingdom  \n468300        0.0         NaN  United Kingdom  \n468302        0.0         NaN  United Kingdom  \n469110        0.0         NaN  United Kingdom  \n472222        0.0         NaN  United Kingdom  \n473440        0.0         NaN  United Kingdom  \n478932        0.0         NaN  United Kingdom  \n478933        0.0         NaN  United Kingdom  \n492198        0.0         NaN  United Kingdom  \n497569        0.0         NaN  United Kingdom  \n500145        0.0         NaN  United Kingdom  \n514287        0.0         NaN  United Kingdom  \n514647        0.0         NaN  United Kingdom  \n517719        0.0         NaN  United Kingdom  \n522162        0.0         NaN  United Kingdom  \n524472        0.0         NaN  United Kingdom  \n535322        0.0         NaN  United Kingdom  \n535331        0.0         NaN  United Kingdom  \n\n[433 rows x 8 columns]\n"}]},{"cell_type":"code","source":"#remove negative quantity\nnegative_quantity = np.where(dataframe_outliers_removed[\"Quantity\"] < 0)\nnegative_quantity = list(negative_quantity[0])\ndataframe_outliers_removed.drop(dataframe_outliers_removed.index[negative_quantity], inplace = True)","metadata":{},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"fig5, ax5 = plt.subplots()\ngreen_diamond = dict(markerfacecolor='g', marker='D')\nax5.set_title('After Removing Outliers Box plot of Quantity')\ndataframe_outliers_removed.boxplot(column=[\"Quantity\"],flierprops=green_diamond)\nplt.show()","metadata":{"scrolled":true},"execution_count":13,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e429ceb8>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"Your explanation:\n- After plotting a boxplot, it is certain that there are outliers in both Quantity and UnitPrice. The outliers are removed based on IQR method and plotted subsequently are the boxplots of the two. \n- It can be observed that the quantity contains negative values. After exploring the dataset, together with that both CustomerID, UnitPrice are unavailible. Description is availible for partial dataset. So it is clear that the dataset does not represent purchases.Therefore it is also removed.","metadata":{"colab_type":"text","id":"GEMVC1cfJUec"}},{"cell_type":"markdown","source":"# Visualization","metadata":{"colab_type":"text","id":"jxkDUfn3Lehi"}},{"cell_type":"markdown","source":"### Question 4 (Stream graph)\nCreate a stream graph that visualizes the number  of purchases (invoices) per country over time.\n\n4.(a) Modify the data set to only contain purchases made in the countries Belgium, Ireland (EIRE), France, Germany, the Netherlands, Norway, Portugal, Spain and Switzerland.","metadata":{"colab_type":"text","id":"eTMS8jZIL2s_"}},{"cell_type":"code","source":"dataframe_for_stream_graph=dataframe_outliers_removed\ndataframe_for_stream_graph=dataframe_for_stream_graph.loc[dataframe_for_stream_graph['Country'].isin(['Belgium','EIRE','France','Netherlands','Germany','Norway','Portugal','Spain','Switzerland'])]","metadata":{"colab":{},"colab_type":"code","id":"Ij3QO-Y_MR3A"},"execution_count":14,"outputs":[]},{"cell_type":"markdown","source":"4.(b) Modify the data set such that it shows per month for each country how many purchases were made (i.e. how many invoices were created).","metadata":{"colab_type":"text","id":"jQTj0TziMUBu"}},{"cell_type":"code","source":"\ndataframe_for_stream_graph['InvoiceDate'] = pd.to_datetime(dataframe_for_stream_graph['InvoiceDate'], errors='coerce')\n    \ndataframe_for_stream_graph['InvoiceDate'] = dataframe_for_stream_graph['InvoiceDate'].dt.strftime('%Y-%m')\n\ndf_groupby = dataframe_for_stream_graph.groupby(['Country','InvoiceDate']).size()\n\ndf_groupby = df_groupby.to_frame().reset_index()\n\ndf_groupby.columns = ['Country', 'InvoiceDate', 'Frequency']\n\ncountry_list=['Belgium',' EIRE', 'France', 'Netherlands', 'Germany','Norway','Portugal','Spain','Switzerland']","metadata":{"colab":{},"colab_type":"code","id":"ADMbVAjOMY-7"},"execution_count":15,"outputs":[]},{"cell_type":"markdown","source":"4.(c) Use the modified data to create a stream graph. ","metadata":{"colab_type":"text","id":"odTJGKrRMcwG"}},{"cell_type":"code","source":"invoiceDate = df_groupby['InvoiceDate'].unique().tolist()\n\nBelgium, EIRE, France, Germany, Netherlands, Norway, Portugal, Spain, Switzerland = ([] for i in range(9))\n\nfor i in range(0, len(df_groupby)):\n    if(df_groupby['Country'][i] == \"Belgium\"):\n        Belgium.append(df_groupby['Frequency'][i])\n    \n    if(df_groupby['Country'][i] == \"EIRE\"):\n        EIRE.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"France\"):\n        France.append(df_groupby['Frequency'][i])\n    \n    if(df_groupby['Country'][i] == \"Germany\"):\n        Germany.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"Netherlands\"):\n        Netherlands.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"Norway\"):\n        Norway.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"Portugal\"):\n        Portugal.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"Spain\"):\n        Spain.append(df_groupby['Frequency'][i])\n        \n    if(df_groupby['Country'][i] == \"Switzerland\"):\n        Switzerland.append(df_groupby['Frequency'][i])\n    \nNorway.insert(1, 0)\nNorway.insert(4, 0)\nNorway.insert(5, 0)\nSwitzerland.append(0)\n\n\n#stacking our values vertically\ny = np.vstack(country_list)\n\n#print(country_list)\n\nfig, ax = plt.subplots(figsize=(20, 10))\n#modifying the axis\nax.stackplot(invoiceDate, Belgium, EIRE, France, Germany, Netherlands, Norway, Portugal, Spain, Switzerland, labels=country_list,baseline='wiggle')\nax.legend(loc='upper left')\nplt.show()","metadata":{"colab":{},"colab_type":"code","id":"V34Z_doDMg7W","scrolled":true},"execution_count":16,"outputs":[{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e432ccc0>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"4.(d) Use this graph to compare the purchases made by each country. Can you find\ninteresting differences between the countries? Are there certain times were the\nsales were particularly high/ particularly low?","metadata":{"colab_type":"text","id":"XXdSd-59Miy8"}},{"cell_type":"markdown","source":"- Belgium: Purchases made in Belgium in the period shows a downward trend overall. There was an rise in purchases in months 02-2011,04-2011,07-2011.There was an drop in purchases in months 01-2011,03-2011,11-2011. The purchases dropped very low in the month 11-2011 and then a step rise in purchases can be observed in 12-2011.\n- EIRE: Purchases made in EIRE in the period shows a downward trend overall. There was an rise in purchases in months 02-2011,04-2011,07-2011.There was an drop in purchases in months 01-2011,03-2011,11-2011. The purchases dropped very low in the month 11-2011 and then a step rise in purchases can be observed in 12-2011.\n- France: Purchases made in France in the period shows a downward trend overall. There was an rise in purchases in months 02-2011,04-2011,09-2011.There was an drop in purchases in months 01-2011,03-2011,11-2011. The purchases dropped very low in the month 11-2011 and then a step rise in purchases can be observed in 12-2011.\n- Netherlands: Purchases made in Netherlands in the period shows a constant trend overall. There was an rise in purchases in months 05-2011.There was an drop in purchases in months 01-2011,09-2011. The purchases dropped very low in the month 10-2011 and then a step rise in purchases can be observed in 11-2011.\n- Germany: Purchases made in Germany in the period shows a constant trend overall. There was an rise in purchases in months 01-2011,03-2011,05-2011.There was an drop in purchases in months 02-2011,06-2011. The purchases rose high in the month 11-2011 and then a steep drop in purchases can be observed in 12-2011.\n- Norway:Purchases made in Norway in the period shows a constant-growing trend overall. The purchases rose high in the month 11-2011 and then a steep drop in purchases can be observed in 12-2011.\n- Portugal:Purchases made in Portugal in the period shows a constant-growing trend overall.There was a rise in purchases in the months 10-2011 and 11-2011 and then a step drop in purchases can be observed in 12-2011.\n- Spain:Purchases made in Spain in the period shows a growth trend overall. There was an rise in purchases in months 01-2011,03-2011,06-2011,10-2011,11-2011.The purchases dropped, steep decrease in purchases can be observed in 12-2011.\n- Switzerland:Purchases made in Switzerland in the period shows a growth trend overall. There was an rise in purchases in months 01-2011,03-2011,06-2011,10-2011,11-2011.Then a steep drop in purchases can be observed in 12-2011.\n\n### <B> Overall trends:<B>\n    - France, EIRE and Belgium have similar trends. The purchases are on the decreasing trend in all three countries. They  have all decreasing purchases in months of January, March and November. In December 2011 a steep rise is seen.\n   \n    -Netherlands, Germany,Norway have a constant trend. November 2011 is having steep rise and then a drop in purchase can    be observed in Dec 2011\n    \n    -Portugal,Spain and Netherlands have growth trend throughout the year. Steep growth can be seen during the November 2011 and then a steep decrease is seen by December 2011\n","metadata":{"colab_type":"text","id":"SutofybpMrKN"}},{"cell_type":"markdown","source":"### Question 5 (Heat map)\nCreate a heat map that visualizes how much (in sterling) each country purchases per month. \n\n5.(a) Modify the data set to only contain purchases made in the countries Belgium, Ireland (EIRE), France, Germany, the Netherlands, Norway, Portugal, Spain and Switzerland. (Or use the version of the data set that you created for question 4 a).","metadata":{"colab_type":"text","id":"jkdo7SE6Mw3N"}},{"cell_type":"code","source":"dataframe_for_stream_graph2=dataframe_outliers_removed\n\ndataframe_for_stream_graph2=dataframe_for_stream_graph2.loc[dataframe_for_stream_graph2['Country'].isin(['Belgium','EIRE','France','Netherlands','Germany','Norway','Portugal','Spain','Switzerland'])]\n","metadata":{"colab":{},"colab_type":"code","id":"5yVXnl6WNZ3N"},"execution_count":17,"outputs":[]},{"cell_type":"markdown","source":"5.(b) Modify the data set such that it shows per month how much money (in sterling) was spent in the shop per country.","metadata":{"colab_type":"text","id":"cEV9doRYNanl"}},{"cell_type":"code","source":"dataframe_for_stream_graph['InvoiceDate'] = pd.to_datetime(dataframe_for_stream_graph['InvoiceDate'], errors='coerce')\n    \ndataframe_for_stream_graph['InvoiceDate'] = dataframe_for_stream_graph['InvoiceDate'].dt.strftime('%Y-%m')\n\ndf_groupby = dataframe_for_stream_graph.groupby(['Country','InvoiceDate']).size()\n\ndf_groupby = df_groupby.to_frame().reset_index()\n\ndf_groupby.columns = ['Country', 'InvoiceDate', 'Frequency']\n\n\ndf_groupby1= dataframe_for_stream_graph.groupby(['Country', 'InvoiceDate'])['UnitPrice'].sum()\n\ndf_groupby1 = df_groupby1.to_frame().reset_index()\n\ndf_groupby1.columns = ['Country', 'InvoiceDate', 'UnitPrice']\n\n\ndf_groupby['UnitPrice']=df_groupby1['UnitPrice']\n\ndf_groupby['Sales']=df_groupby['Frequency']*df_groupby1['UnitPrice']\n","metadata":{"colab":{},"colab_type":"code","id":"WdKt5vIGNfAF"},"execution_count":18,"outputs":[]},{"cell_type":"markdown","source":"5.(c) Use the modified data to create a heat map. ","metadata":{"colab_type":"text","id":"Z9OxLF1xNg6d"}},{"cell_type":"code","source":"f, ax = plt.subplots(figsize=(10, 10))\npivotted=df_groupby.pivot('Country','InvoiceDate','Sales')\nsns.heatmap(pivotted,linewidths=1.5, ax=ax)","metadata":{"colab":{},"colab_type":"code","id":"BeqD_H9BNsWd"},"execution_count":19,"outputs":[{"execution_count":19,"output_type":"execute_result","data":{"text/plain":"<matplotlib.axes._subplots.AxesSubplot at 0x182e4306c88>"},"metadata":{}},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e68774e0>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"5.(d) Compare the amount of the purchases over time and between each country. ","metadata":{"colab_type":"text","id":"K_DdfvB7NuH1"}},{"cell_type":"markdown","source":"- Belgium: The amount purchased in Belgium during the period on average is less than 600k sterlings. \n\n- EIRE: The amount of purchased in EIRE during the period on average is between 300k-6M. Months September,October, November has seen the highest purchase amounts.\n\n- France: The amount of purchased in France during the period on average is between 600k-1.8M. Months September,October, November has seen the highest purchase amounts.\n\n- Netherlands: The amount of purchased in Netherlands during the period on average is less than 600k \n\n- Germany: The amount of purchased in Germany during the period is increasing every month from Dec 2010(from 600k-2.4M). A low amount is seen in Dec 2011. \n\n- Norway:The amount of purchased in Norway during the period on average is less than 600K \n\n- Portugal:The amount of purchased in Netherlands during the period on average is less than 600k.\n\n- Spain:The amount of purchased in Spain ranges 600K-1.2M.\n\n- Switzerland:The amount of purchased in Switzerland ranges 0-600k.\n\n","metadata":{"colab_type":"text","id":"nYBMrbELNz9e"}},{"cell_type":"markdown","source":"### Question 6 (Interpretation)\nCompare the results obtained from the stream graph and the heat map. Is there a relation between the number of purchases and the amount purchased in sterling?\n","metadata":{"colab_type":"text","id":"y3OJYgdqN38Y"}},{"cell_type":"markdown","source":"- There is no definite relationship between purchases and the amount purchased in sterling.\n- Countries such as Germany,Norway,Spain with the increase in number of purchases there is an increase in amount\n- Countries Belgium, Netherlands,Portugal,Switzerland have almost constant amount spent even with increase or decrease in purchases\n- Countries EIRE,France has shown higher amount purchased even with lower number of purchases","metadata":{"colab_type":"text","id":"WSReYGIFN_c9"}},{"cell_type":"markdown","source":"# Clustering\nPresume that the business analyst would like to cluster transactions with similar types of products into the same group (here don’t consider the quantity of the products). For each product, only use its ‘StockCode’ to represent it. All the results here should be based on the preprocessed data set obtained from question 1 to 3 of this assignment. Presume that this obtained data set from question 1 to 3 has a variable name ‘cluster_dataset’ and is expressed by Pandas DataFrame in your code.\n\n### Question 7 (Data transformation and clustering)\n7.(a) Calculate and show the number of occurrences of each product in data set   ‘cluster_dataset’. For example, if a product appears in a transaction, then its occurrence number will be increased by 1 (do not consider the quantity of this product here). Preserve the 100 most frequent products and remove all the other products in ‘cluster_dataset’. For example, if a row in ‘cluster_dataset’ contains unqualified product, then remove this row from ‘cluster_dataset’. Show the new ‘cluster_dataset’ in your result.\n","metadata":{"colab_type":"text","id":"Zfz0fzTpA-3j"}},{"cell_type":"code","source":"cluster_dataset=dataframe_outliers_removed\n\nto_sort=cluster_dataset['StockCode'].value_counts().to_dict()\n\nto_be_removed=list(to_sort.keys())[101:]\n\ncluster_dataset = cluster_dataset[~cluster_dataset['StockCode'].isin(to_be_removed)]\n\ncluster_dataset.head()\n","metadata":{"colab":{},"colab_type":"code","id":"tV-Zzxo-BJwS"},"execution_count":20,"outputs":[{"execution_count":20,"output_type":"execute_result","data":{"text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>InvoiceNo</th>\n      <th>StockCode</th>\n      <th>Description</th>\n      <th>Quantity</th>\n      <th>InvoiceDate</th>\n      <th>UnitPrice</th>\n      <th>CustomerID</th>\n      <th>Country</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>536365</td>\n      <td>85123A</td>\n      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n      <td>6</td>\n      <td>2010-12-01 08:26:00</td>\n      <td>2.55</td>\n      <td>17850.0</td>\n      <td>United Kingdom</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>536367</td>\n      <td>21754</td>\n      <td>HOME BUILDING BLOCK WORD</td>\n      <td>3</td>\n      <td>2010-12-01 08:34:00</td>\n      <td>5.95</td>\n      <td>13047.0</td>\n      <td>United Kingdom</td>\n    </tr>\n    <tr>\n      <th>21</th>\n      <td>536368</td>\n      <td>22960</td>\n      <td>JAM MAKING SET WITH JARS</td>\n      <td>6</td>\n      <td>2010-12-01 08:34:00</td>\n      <td>4.25</td>\n      <td>13047.0</td>\n      <td>United Kingdom</td>\n    </tr>\n    <tr>\n      <th>28</th>\n      <td>536370</td>\n      <td>22726</td>\n      <td>ALARM CLOCK BAKELIKE GREEN</td>\n      <td>12</td>\n      <td>2010-12-01 08:45:00</td>\n      <td>3.75</td>\n      <td>12583.0</td>\n      <td>France</td>\n    </tr>\n    <tr>\n      <th>49</th>\n      <td>536373</td>\n      <td>85123A</td>\n      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n      <td>6</td>\n      <td>2010-12-01 09:02:00</td>\n      <td>2.55</td>\n      <td>17850.0</td>\n      <td>United Kingdom</td>\n    </tr>\n  </tbody>\n</table>\n</div>","text/plain":"   InvoiceNo StockCode                         Description  Quantity  \\\n0     536365    85123A  WHITE HANGING HEART T-LIGHT HOLDER         6   \n17    536367     21754            HOME BUILDING BLOCK WORD         3   \n21    536368     22960            JAM MAKING SET WITH JARS         6   \n28    536370     22726          ALARM CLOCK BAKELIKE GREEN        12   \n49    536373    85123A  WHITE HANGING HEART T-LIGHT HOLDER         6   \n\n           InvoiceDate  UnitPrice  CustomerID         Country  \n0  2010-12-01 08:26:00       2.55     17850.0  United Kingdom  \n17 2010-12-01 08:34:00       5.95     13047.0  United Kingdom  \n21 2010-12-01 08:34:00       4.25     13047.0  United Kingdom  \n28 2010-12-01 08:45:00       3.75     12583.0          France  \n49 2010-12-01 09:02:00       2.55     17850.0  United Kingdom  "},"metadata":{}}]},{"cell_type":"markdown","source":"7.(b) Based on question a, please reorganize the data from ‘cluster_dataset’ and generate a new data set ‘cluster_dataset_new’ which has a suitable format (for k-means) for solving the transaction clustering problem mentioned above. Show the data from ‘cluster_dataset_new’ by using Pandas DataFrame in your result, where the index should be consistent with the values of 'InvoiceNo', the column name should be consistent with the values of 'StockCode' and each element in this DataFrame should have a value 0 or 1.","metadata":{"colab_type":"text","id":"zvM5UjQclH0j"}},{"cell_type":"code","source":"cluster_dataset.StockCode = cluster_dataset.StockCode.astype(str)\n\ncluster_dataset_new = pd.crosstab(cluster_dataset.InvoiceNo, cluster_dataset.StockCode)\n\ncluster_dataset_new.head()","metadata":{"colab":{},"colab_type":"code","id":"VMLEa0gslXcD"},"execution_count":21,"outputs":[{"execution_count":21,"output_type":"execute_result","data":{"text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th>StockCode</th>\n      <th>20712</th>\n      <th>20713</th>\n      <th>20719</th>\n      <th>20723</th>\n      <th>20724</th>\n      <th>20725</th>\n      <th>20726</th>\n      <th>20727</th>\n      <th>20728</th>\n      <th>20914</th>\n      <th>...</th>\n      <th>82583</th>\n      <th>84378</th>\n      <th>84380</th>\n      <th>84879</th>\n      <th>84946</th>\n      <th>85099B</th>\n      <th>85099C</th>\n      <th>85099F</th>\n      <th>85123A</th>\n      <th>85152</th>\n    </tr>\n    <tr>\n      <th>InvoiceNo</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>536365</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>536367</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>536368</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>536370</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>536373</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 101 columns</p>\n</div>","text/plain":"StockCode  20712  20713  20719  20723  20724  20725  20726  20727  20728  \\\nInvoiceNo                                                                  \n536365         0      0      0      0      0      0      0      0      0   \n536367         0      0      0      0      0      0      0      0      0   \n536368         0      0      0      0      0      0      0      0      0   \n536370         0      0      0      0      0      0      0      0      0   \n536373         0      0      0      0      0      0      0      0      0   \n\nStockCode  20914  ...    82583  84378  84380  84879  84946  85099B  85099C  \\\nInvoiceNo         ...                                                        \n536365         0  ...        0      0      0      0      0       0       0   \n536367         0  ...        0      0      0      0      0       0       0   \n536368         0  ...        0      0      0      0      0       0       0   \n536370         0  ...        0      0      0      0      0       0       0   \n536373         0  ...        0      0      0      0      0       0       0   \n\nStockCode  85099F  85123A  85152  \nInvoiceNo                         \n536365          0       1      0  \n536367          0       0      0  \n536368          0       0      0  \n536370          0       0      0  \n536373          0       1      0  \n\n[5 rows x 101 columns]"},"metadata":{}}]},{"cell_type":"markdown","source":"7.(c) Try values 2, 3, 4, 5 for parameter 'n_clusters' for the k-means function from Scikit-Learn over the data set ‘cluster_dataset_new’ generated in question b. Show the ‘within cluster variation’ (also called ‘sum of squared distances’) of the generated clusters for each different setting for ‘n_clusters’ in your result. Also write down the value that you have tried for setting 'n_clusters' which can help generate the best clustering results and explain how you make this decision.","metadata":{"colab_type":"text","id":"ox2r6Nvxlcm4"}},{"cell_type":"code","source":"import collections\nfrom sklearn.cluster import KMeans\nfrom sklearn.metrics.pairwise import euclidean_distances\nfrom sklearn.metrics.pairwise import pairwise_distances\nfrom scipy.sparse import lil_matrix\nfrom scipy.spatial.distance import cdist\nfrom sklearn.metrics import silhouette_score\nfrom scipy.sparse import csr_matrix\n\ndataframe_collection = {} \n\nsum_of_square_distance_array = []\n\nfor k in range (2, 6):\n    \n    kmeans_model = KMeans(n_clusters=k, random_state=1).fit(cluster_dataset_new)\n    labels = kmeans_model.labels_\n    interia = kmeans_model.inertia_\n    sum_of_square_distance_array.append(interia)\n    print (\"k:\",k, \" Inertia:\", interia)\n    centroids = kmeans_model.cluster_centers_\n    dists = euclidean_distances(centroids)\n    print (\"Mean euclidean distances:\", dists.mean())\n    cluster_map = pd.DataFrame()\n    cluster_map['data_index'] = cluster_dataset_new.index.values\n    cluster_map['cluster'] = labels\n    dataframe_collection[k] = cluster_map\n    print (\"\\n\")\n    \nK = range(2,6)\nplt.plot(K, sum_of_square_distance_array, 'bx-')\nplt.xlabel('K')\nplt.ylabel('Sum_of_squared_distances')\nplt.title('Elbow method for Optimal K')\nplt.show()","metadata":{"colab":{},"colab_type":"code","id":"Q68MWZDIlrOn"},"execution_count":28,"outputs":[{"name":"stdout","output_type":"stream","text":"k: 2  Inertia: 79973.55909271068\nMean euclidean distances: 0.9272201988285486\n\n\nk: 3  Inertia: 77739.17802918644\nMean euclidean distances: 1.4108106459869036\n\n\nk: 4  Inertia: 75903.60173950238\nMean euclidean distances: 1.7464549897400605\n\n\nk: 5  Inertia: 74515.68672767424\nMean euclidean distances: 1.7920839754413749\n\n\n"},{"output_type":"display_data","data":{"image/png":"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\n","text/plain":"<matplotlib.figure.Figure at 0x182e48a1128>"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"Your explanation:\n- From the plotted graph of sum of squared distances vs K, it can be observed that, the value of inertia decreases as the number of clusters increase.\n- I am choosing 4 as my optimal K since from the plotted graph, the elbow point in the inertia graph is a good choice because after that the change in the value of inertia isn’t significant. (If the plot looks like an arm, then the elbow on the arm is optimal k.)","metadata":{"colab_type":"text","id":"q_X_6ev3Ojqu"}},{"cell_type":"markdown","source":"# Frequent Itemsets and Association Rules\n For the clusters output by k-means function with the best 'n_clusters' from question 7, the business analyst now would like to research on the frequent purchase behaviours and specific purchase rules for each cluster.\n### Question 8 (Data transformation and mining frequent itemsets and association rules)\n8.(a) Set the minimum support for finding the frequent purchase behaviours to 0.2. Please provide the business analyst with the qualified purchase behaviours. For each product, only use its ‘StockCode’ to represent it. Also show the data set prepared for each cluster for mining the frequent behaviours by using Pandas DataFrame in your result, the data set for the cluster k should have the variable name 'fpb_data_k' in your code.","metadata":{"colab_type":"text","id":"ekfbWzxdBKU8"}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"8.(b) Furthermore, the business analyst would like to analyze the purchase behaviour of the citizens from ‘United Kingdom’ for each cluster. Specifically speaking, he wants to discover if there exist some rules which indicate that the citizens from ‘United Kingdom’ tend to buy some specific products for each cluster. Set the minimum support to 0.2, minimum confidence to 0.7. Please discover and show such rules (only show the rules with ‘United Kingdom’ appearing in antecedents in the rules) for each cluster for the business analyst. Also show the data sets prepared for each cluster for mining the relevant rules by using Pandas DataFrame in you result, the data set for cluster k should have the variable name 'r_data_k' in your code.","metadata":{"colab_type":"text","id":"908qPmK3Oyzf"}},{"cell_type":"code","source":"# your code","metadata":{"colab":{},"colab_type":"code","id":"246Z21sPnlRV"},"execution_count":61,"outputs":[]},{"cell_type":"markdown","source":"# Text Mining\n### Question 12 (Model based on binary document-term matrix)\nPerform preprocessing on the corpus (all lowercase, no punctuation, tokenization, stemming, stopword removal) and obtain a binary document-term matrix; train a logistic classifier.","metadata":{"colab_type":"text","id":"G9-Gqj3qBXjw"}},{"cell_type":"code","source":"from sklearn.feature_extraction.text import CountVectorizer\nimport nltk\nfrom nltk.tokenize import sent_tokenize, word_tokenize\nimport nltk.data\nfrom nltk.corpus import stopwords\nimport string\nfrom nltk.stem.snowball import SnowballStemmer\nfrom sklearn.linear_model import LogisticRegression,SGDClassifier\nfrom sklearn.metrics import roc_curve, roc_auc_score, auc\nfrom sklearn.linear_model import SGDClassifier\nfrom sklearn.pipeline import Pipeline\nimport numpy as np\n\ntext_train = pd.read_csv(\"pg_train.csv\",sep='delimiter',header=None)\ntext_test=pd.read_csv(\"pg_test.csv\",sep='delimiter',header=None)\n\n#Removing '#' and making columns\ntext_train = text_train.rename(columns={text_train.columns[0]: \"colname\"})\ntext_train = text_train.colname.str.split(\"#\", expand=True)\n\n\ntext_train.rename(columns={ text_train.columns[0]: \"target\" }, inplace = True)\ntext_train.rename(columns={ text_train.columns[1]: \"text\" }, inplace = True)\n\n#Converting to lower case\ntext_lower_case=text_train[\"text\"][:].str.lower().to_frame()\ntext_train[\"text\"]=text_lower_case[\"text\"]\n\n# Remove punctuation \ntext_train[\"text\"]=text_train[\"text\"].str.replace('[{}]'.format(string.punctuation), '')\n\n\n#tokenizer\ntext_train[\"text\"] = text_train.apply(lambda row: nltk.word_tokenize(row['text']), axis=1)\n\n\n#Stemming\nstemmer = SnowballStemmer(\"english\", ignore_stopwords=True)\ntext_train[\"text\"] = text_train[\"text\"].apply(lambda x: [stemmer.stem(y) for y in x])\n\n\n#Stopword removal\nstoplist = set(stopwords.words('english'))\ntext_train[\"text\"]=text_train[\"text\"].apply(lambda x: [item for item in x if item not in stoplist])\n\n#Document Matrix\nvect = CountVectorizer(tokenizer=lambda doc:doc, analyzer='word',binary=True,lowercase=False)\ndocument_term_matrix=vect.fit_transform(text_train.text,y=None)\n\ntext_classification_binary = Pipeline([('vect', vect), ('logistic', SGDClassifier(loss='hinge', penalty='l2',alpha=1e-3, max_iter = 30, random_state = 42))])\ntext_classification_binary = text_classification_binary.fit(text_train.text, text_train.target)\n","metadata":{"colab":{},"colab_type":"code","id":"t1JdbpAJBTHQ"},"execution_count":62,"outputs":[]},{"cell_type":"markdown","source":"### Question 13 (Model based on document-term matrix of counts)\nPerform preprocessing on the corpus (all lowercase, no punctuation, tokenization, stemming, stopword removal) and obtain a document-term matrix of counts; train a logistic classifier.\n","metadata":{"colab_type":"text","id":"HFoouarHPWfK"}},{"cell_type":"code","source":"#Document Matrix\nvect = CountVectorizer(tokenizer=lambda doc:doc, analyzer='word',binary=False,lowercase=False)\ndocument_term_matrix=vect.fit_transform(text_train.text,y=None)\n\n\ntext_classification = Pipeline([('vect', vect), ('logistic', SGDClassifier(loss='hinge', penalty='l2',alpha=1e-3, max_iter = 30, random_state = 42))])\ntext_classification = text_classification.fit(text_train.text, text_train.target)\n","metadata":{"colab":{},"colab_type":"code","id":"vDjnCPi2BaNm"},"execution_count":63,"outputs":[]},{"cell_type":"markdown","source":"### Question 14 (Model based on tf-idf document-term matrix)\nPerform preprocessing on the corpus (all lowercase, no punctuation, tokenization, stemming, stopword removal) and obtain a tf-idf scores document-term matrix; train a logistic classifier.\n","metadata":{"colab_type":"text","id":"SpcgbGTOPp3A"}},{"cell_type":"code","source":"from sklearn.feature_extraction.text import TfidfTransformer\n\nvect = CountVectorizer(tokenizer=lambda doc:doc, analyzer='word',binary=False,lowercase=False)\ndocument_term_matrix=vect.fit_transform(text_train.text,y=None)\n\ntfidf_transformer = TfidfTransformer()\nX_train_tfidf = tfidf_transformer.fit_transform(document_term_matrix)\n\n\nfrom sklearn.linear_model import SGDClassifier\n\ntext_clf = Pipeline([('vect', vect), ('tf-idf', TfidfTransformer()), ('Logistic', SGDClassifier(loss='hinge', penalty='l2',alpha=1e-3, max_iter=30, random_state=42))])\n\ntext_clf = text_clf.fit(text_train.text, text_train.target)\n\n","metadata":{"colab":{},"colab_type":"code","id":"1FJgzMKlPus6"},"execution_count":64,"outputs":[]},{"cell_type":"markdown","source":"### Question 15 (Model based on doc2vec)\nPerform preprocessing on the corpus (all lowercase, no punctuation, tokenization, stemming, stopword removal) and obtain a doc2vec embedding in order to reduce the dimension of the document vector to 300; use the doc2vec model you just trained to convert the training set to a set of document vectors; train a logistic classifier.\n","metadata":{"colab_type":"text","id":"4IV9XRbKPwGW"}},{"cell_type":"code","source":"from tqdm import tqdm\ntqdm.pandas(desc=\"progress-bar\")\nfrom gensim.models import Doc2Vec\nfrom sklearn import utils\nfrom sklearn.model_selection import train_test_split\nimport gensim\nfrom sklearn.linear_model import LogisticRegression\nfrom gensim.models.doc2vec import TaggedDocument\nimport re\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom gensim.test.utils import common_texts\nfrom gensim.models.doc2vec import Doc2Vec, TaggedDocument\nfrom bs4 import BeautifulSoup\n\nimport multiprocessing\ncores = multiprocessing.cpu_count()\n\ndf_train = pd.read_csv(\"pg_train.csv\",sep='delimiter',header=None)\ndf_test=pd.read_csv(\"pg_test.csv\",sep='delimiter',header=None)\n\ndf_train = df_train.rename(columns={df_train.columns[0]: \"colname\"})\ndf_train = df_train.colname.str.split(\"#\", expand=True)\n\ndf_test = df_test.rename(columns={df_test.columns[0]: \"colname\"})\ndf_test = df_test.colname.str.split(\"#\", expand=True)\n\ndf_train.rename(columns={ df_train.columns[0]: \"target\" }, inplace = True)\ndf_train.rename(columns={ df_train.columns[1]: \"text\" }, inplace = True)\ndf_train = df_train[pd.notnull(df_train['text'])]\n\ndf_test.rename(columns={ df_test.columns[0]: \"target\" }, inplace = True)\ndf_test.rename(columns={ df_test.columns[1]: \"text\" }, inplace = True)\ndf_test = df_test[pd.notnull(df_test['text'])]\n\n\ndf_train.index = range(4499)\ndf_train['text'].apply(lambda x: len(x.split(' '))).sum() #572257 words\n\ndf_train.index = range(4499)\ndf_train['text'].apply(lambda x: len(x.split(' '))).sum() #572257 words\n","metadata":{"colab":{},"colab_type":"code","id":"cv8oGoxlQCzm"},"execution_count":65,"outputs":[{"execution_count":65,"output_type":"execute_result","data":{"text/plain":"572257"},"metadata":{}}]},{"cell_type":"code","source":"#Text Preprocessing\ndef cleanText(text):\n    text = BeautifulSoup(text, \"lxml\").text\n    text = re.sub(r'\\|\\|\\|', r' ', text) \n    text = re.sub(r'http\\S+', r'<URL>', text)\n    text = text.lower()\n    text = text.replace('x', '')\n    return text\ndf_train['text'] = df_train['text'].apply(cleanText)\ndf_test['text'] = df_test['text'].apply(cleanText)","metadata":{},"execution_count":66,"outputs":[]},{"cell_type":"code","source":"#Stopword removal and tokenize\nimport nltk\nfrom nltk.corpus import stopwords\ndef tokenize_text(text):\n    tokens = []\n    for sent in nltk.sent_tokenize(text):\n        for word in nltk.word_tokenize(sent):\n            if len(word) < 2:\n                continue\n            tokens.append(word.lower())\n    return tokens\n\n#creating tagged document for Doc2vec\ntrain_tagged = df_train.apply(lambda r: TaggedDocument(words=tokenize_text(r['text']), tags=[r.target]), axis=1)\ntest_tagged = df_test.apply(lambda r: TaggedDocument(words=tokenize_text(r['text']), tags=[r.target]), axis=1)","metadata":{},"execution_count":67,"outputs":[]},{"cell_type":"code","source":"model_dbow = Doc2Vec(dm=0, vector_size=300, negative=5, hs=0, min_count=2, sample = 0, workers=cores)\nmodel_dbow.build_vocab([x for x in tqdm(train_tagged.values)])\n","metadata":{},"execution_count":68,"outputs":[{"name":"stderr","output_type":"stream","text":"100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900638.30it/s]\n"}]},{"cell_type":"code","source":"for epoch in range(30):\n    model_dbow.train(utils.shuffle([x for x in tqdm(train_tagged.values)]), total_examples=len(train_tagged.values), epochs=1)\n    model_dbow.alpha -= 0.002\n    model_dbow.min_alpha = model_dbow.alpha","metadata":{},"execution_count":69,"outputs":[{"name":"stderr","output_type":"stream","text":"100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 749917.49it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900337.50it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900337.50it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1501207.14it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1125166.88it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 643067.53it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900165.71it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1125502.43it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900165.71it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900337.50it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900337.50it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 643199.05it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900079.83it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1125502.43it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900423.42it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 899951.05it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1125435.30it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 750275.29it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1124764.48it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 750305.12it/s]\n100%|█████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 1125166.88it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900165.71it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900552.34it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900337.50it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 750245.46it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900165.71it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 643264.83it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 750126.16it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 562667.31it/s]\n100%|██████████████████████████████████████████████████████████████████████████| 4499/4499 [00:00<00:00, 900380.46it/s]\n"}]},{"cell_type":"code","source":"def vec_for_learning(model, tagged_docs):\n    sents = tagged_docs.values\n    targets, regressors = zip(*[(doc.tags[0], model.infer_vector(doc.words, steps=20)) for doc in sents])\n    return targets, regressors","metadata":{},"execution_count":86,"outputs":[]},{"cell_type":"code","source":"y_train, X_train = vec_for_learning(model_dbow, train_tagged)\ny_test, X_test = vec_for_learning(model_dbow, test_tagged)\nlogreg = LogisticRegression(n_jobs=1, C=1e5)\nlogreg.fit(X_train, y_train)\n","metadata":{},"execution_count":87,"outputs":[{"execution_count":87,"output_type":"execute_result","data":{"text/plain":"LogisticRegression(C=100000.0, class_weight=None, dual=False,\n          fit_intercept=True, intercept_scaling=1, max_iter=100,\n          multi_class='warn', n_jobs=1, penalty='l2', random_state=None,\n          solver='warn', tol=0.0001, verbose=0, warm_start=False)"},"metadata":{}}]},{"cell_type":"markdown","source":"### Question 16 (Evaluation)\n16.(a) Predict the classification with the four models on the test data.\n","metadata":{"colab_type":"text","id":"nUTK6JL6QKxX"}},{"cell_type":"code","source":"#Preprocessing Text data\ntext_test=pd.read_csv(\"pg_test.csv\",sep='delimiter',header=None)\n\n#Removing '#' and making columns\ntext_test = text_test.rename(columns={text_test.columns[0]: \"colname\"})\ntext_test = text_test.colname.str.split(\"#\", expand=True)\ntext_test.rename(columns={ text_test.columns[0]: \"target\" }, inplace = True)\ntext_test.rename(columns={ text_test.columns[1]: \"text\" }, inplace = True)\n#Converting to lower case\ntext_lower_case=text_test[\"text\"][:].str.lower().to_frame()\ntext_test[\"text\"]=text_lower_case[\"text\"]\n# Remove punctuation \ntext_test[\"text\"]=text_test[\"text\"].str.replace('[{}]'.format(string.punctuation), '')\n#tokenizer\ntext_test[\"text\"] = text_test.apply(lambda row: nltk.word_tokenize(row['text']), axis=1)\n#Stemming\nstemmer = SnowballStemmer(\"english\", ignore_stopwords=True)\ntext_test[\"text\"] = text_test[\"text\"].apply(lambda x: [stemmer.stem(y) for y in x])\n#Stopword removal\nstoplist = set(stopwords.words('english'))\ntext_test[\"text\"]=text_test[\"text\"].apply(lambda x: [item for item in x if item not in stoplist])","metadata":{"colab":{},"colab_type":"code","id":"HoGSkKTDQSLP"},"execution_count":88,"outputs":[]},{"cell_type":"code","source":"#Predicted classification on Model based on binary document-term matrix\npredicted= text_classification_binary.predict(text_test.text)\nnp.mean(predicted == text_test.target)","metadata":{},"execution_count":89,"outputs":[{"execution_count":89,"output_type":"execute_result","data":{"text/plain":"0.9827242524916944"},"metadata":{}}]},{"cell_type":"code","source":"#Predicted classification on Model based on document-term matrix of counts\npredicted1= text_classification.predict(text_test.text)\nnp.mean(predicted1== text_test.target)","metadata":{},"execution_count":90,"outputs":[{"execution_count":90,"output_type":"execute_result","data":{"text/plain":"0.9720930232558139"},"metadata":{}}]},{"cell_type":"code","source":"#Predicted classification on Model based on tf-idf\npredicted2 =text_clf.predict(text_test.text)\nnp.mean(predicted2== text_test.target)","metadata":{},"execution_count":91,"outputs":[{"execution_count":91,"output_type":"execute_result","data":{"text/plain":"0.987375415282392"},"metadata":{}}]},{"cell_type":"code","source":"#Predicted classification on Model based on doc2vec\ny_pred = logreg.predict(X_test)\nnp.mean(y_pred== df_test.target)","metadata":{},"execution_count":92,"outputs":[{"execution_count":92,"output_type":"execute_result","data":{"text/plain":"0.7089700996677741"},"metadata":{}}]},{"cell_type":"markdown","source":"16.(b) Obtain confusion matrices for the four different models.\n","metadata":{"colab_type":"text","id":"RiB6Dm6KQVY3"}},{"cell_type":"code","source":"from sklearn.metrics import accuracy_score, f1_score, confusion_matrix","metadata":{"colab":{},"colab_type":"code","id":"Vvp_5hEWQZN-"},"execution_count":77,"outputs":[]},{"cell_type":"code","source":"# #Confusion Matrix on Model based on binary document-term matrix\ncm1 = confusion_matrix(text_test.target, predicted)\nprint(cm1)","metadata":{},"execution_count":78,"outputs":[{"name":"stdout","output_type":"stream","text":"[[680   7   2]\n [ 12 498   5]\n [  0   0 301]]\n"}]},{"cell_type":"code","source":"#Confusion Matrix on Model based on document-term matrix of counts\ncm2 = confusion_matrix(text_test.target, predicted1)\nprint(cm2)","metadata":{},"execution_count":79,"outputs":[{"name":"stdout","output_type":"stream","text":"[[673  11   5]\n [ 17 492   6]\n [  1   2 298]]\n"}]},{"cell_type":"code","source":"#Confusion Matrix on Model based on tf-idf\ncm3 = confusion_matrix(text_test.target, predicted2)\nprint(cm3)","metadata":{},"execution_count":80,"outputs":[{"name":"stdout","output_type":"stream","text":"[[680   9   0]\n [  6 507   2]\n [  0   2 299]]\n"}]},{"cell_type":"code","source":"#Confusion Matrix on Model based on doc2vec\ncm = confusion_matrix(y_test, y_pred)\nprint(cm)","metadata":{},"execution_count":81,"outputs":[{"name":"stdout","output_type":"stream","text":"[[564 117   8]\n [273 231  11]\n [ 11   8 282]]\n"}]},{"cell_type":"markdown","source":"16.(c) Obtain accuracy and f1 score for the four different models.\n","metadata":{"colab_type":"text","id":"gi9Xb-JoQagB"}},{"cell_type":"code","source":"# Accuracy and F1 score of Model based on binary document-term matrix\nprint('Testing accuracy %s' % accuracy_score(text_test.target, predicted))\nprint('Testing F1 score: {}'.format(f1_score(text_test.target, predicted, average='weighted')))","metadata":{"colab":{},"colab_type":"code","id":"sem1ajYSQfI_"},"execution_count":82,"outputs":[{"name":"stdout","output_type":"stream","text":"Testing accuracy 0.9827242524916944\nTesting F1 score: 0.9826879683618106\n"}]},{"cell_type":"code","source":"#Accuracy and F1 score of Model based on document-term matrix of counts\nprint('Testing accuracy %s' % accuracy_score(text_test.target, predicted1))\nprint('Testing F1 score: {}'.format(f1_score(text_test.target, predicted1, average='weighted')))","metadata":{},"execution_count":83,"outputs":[{"name":"stdout","output_type":"stream","text":"Testing accuracy 0.9720930232558139\nTesting F1 score: 0.9720531364595433\n"}]},{"cell_type":"code","source":"#Accuracy and F1 score of Model based on tf-idf\nprint('Testing accuracy %s' % accuracy_score(text_test.target, predicted2))\nprint('Testing F1 score: {}'.format(f1_score(text_test.target, predicted2, average='weighted')))","metadata":{},"execution_count":84,"outputs":[{"name":"stdout","output_type":"stream","text":"Testing accuracy 0.987375415282392\nTesting F1 score: 0.9873828743579076\n"}]},{"cell_type":"code","source":"#Accuracy and F1 Score of Model based on doc2vec\nprint('Testing accuracy %s' % accuracy_score(y_test, y_pred))\nprint('Testing F1 score: {}'.format(f1_score(y_test, y_pred, average='weighted')))","metadata":{},"execution_count":85,"outputs":[{"name":"stdout","output_type":"stream","text":"Testing accuracy 0.7156146179401993\nTesting F1 score: 0.7048664079340687\n"}]},{"cell_type":"markdown","source":"16.(d) Briefly comment on the quality of the predictions for the four models.","metadata":{"colab_type":"text","id":"9-sy3DloQg72"}},{"cell_type":"markdown","source":"_Your comment:_\n- TF-IDF, Document-term matrix of counts and Binary Document-term matrix perform much better than doc2vec. All three of these models are able to predict the author based on text with an accuracy greater than 97%.\n- Among them TF-IDF based model performs best and displays best metrics such as accuracy and F1 scores.\n","metadata":{"colab_type":"text","id":"42-LzxBoQn-x"}},{"cell_type":"markdown","source":"# Process Mining\nFor this part, refer to the online docs of pm4py. You will find particularly of interest the documentation on filtering (https://pm4py.github.io/filtering.html, or on the new website http://pm4py.pads.rwth-aachen.de/documentation/filtering-logs/). \nimportant: if you did not do it in the instruction, you should make sure to have the latest pm4py version: to get it is sufficient to type `pip install pm4py --upgrade` from any terminal emulator on Windows  (command prompt, PowerShell, etc) or any terminal on *nix systems. For the details, refer to the study guide and the Process Mining instruction.\n### Question 17 (Trace frequency)\n17.(a) Use the provided event log and identify the least frequent traces and the most frequent traces.\n","metadata":{"colab_type":"text","id":"pGImEFS3BPcJ"}},{"cell_type":"code","source":"import pm4py\nfrom pm4py.objects.log.importer.xes import factory as xes_importer\nfrom pm4py.objects.log.util import sorting\nfrom pm4py.algo.filtering.tracelog.variants import variants_filter\nfrom pm4py.statistics.traces.tracelog import case_statistics\nfrom pm4py.algo.discovery.inductive import factory as inductive_miner\nfrom pm4py.algo.conformance.tokenreplay import factory as token_based_replay_factory\nfrom pm4py.evaluation.replay_fitness import factory as replay_fitness_factory\n\n\n# Importing a XES event log\nlog = xes_importer.import_log('event_log.xes')\nlog = sorting.sort_timestamp(log)\n\nvariants = variants_filter.get_variants(log)\n\nvariants_count = case_statistics.get_variant_statistics(log)\nvariants_count = sorted(variants_count, key=lambda x: x['count'])\nprint(variants_count)\n\nprint(len(log))\n\nmin=variants_count[0]['variant']\nmin1=variants_count[1]['variant']\nmax=variants_count[len(variants_count)-1]['variant']\nmax1=variants_count[len(variants_count)-2]['variant']\n","metadata":{"colab":{},"colab_type":"code","id":"tU_bbK2FBkK0"},"execution_count":58,"outputs":[{"name":"stdout","output_type":"stream","text":"[{'variant': 'Assigned Severity,Under Processing,Waiting,Solves,Closed', 'count': 2251}, {'variant': 'Assigned Severity,Under Processing,Under Processing,Waiting,Solves,Closed', 'count': 2568}, {'variant': 'Assigned Severity,Waiting,Under Processing,Under Processing,Solves,Closed', 'count': 10614}, {'variant': 'Assigned Severity,Waiting,Under Processing,Solves,Closed', 'count': 12188}]\n27621\n"}]},{"cell_type":"markdown","source":"### Question 18 (Process Discovery and Conformance Checking using first filtered event log)\n18.(a) Remove the two least frequent traces and create a new event log out of the original event log without the two least frequent traces.","metadata":{"colab_type":"text","id":"_5_JFMrER2rf"}},{"cell_type":"code","source":"filtered_log = variants_filter.apply(log, [min,min1], parameters={\"positive\": False})\nprint(len(filtered_log))\n","metadata":{"colab":{},"colab_type":"code","id":"UJmxAmghRodA"},"execution_count":59,"outputs":[{"name":"stdout","output_type":"stream","text":"22802\n"}]},{"cell_type":"markdown","source":"18.(b) Use Inductive miner algorithm to discover the process model based on you new event log (the filtered log without two least frequent traces).\n","metadata":{"colab_type":"text","id":"_vjF6dm0RpQX"}},{"cell_type":"code","source":"# Mining for a process tree\n\ntree = inductive_miner.apply_tree(filtered_log)\nnet, initial_marking, final_marking = inductive_miner.apply(filtered_log)\n","metadata":{"colab":{},"colab_type":"code","id":"NqlBAJYkQrWK"},"execution_count":69,"outputs":[]},{"cell_type":"code","source":"# Petri net visualization\nfrom pm4py.visualization.petrinet import factory as pn_vis_factory\n\ngviz_pn = pn_vis_factory.apply(net, initial_marking, final_marking)\npn_vis_factory.view(gviz_pn)","metadata":{},"execution_count":71,"outputs":[{"execution_count":71,"output_type":"execute_result","data":{"image/png":"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\n","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}]},{"cell_type":"markdown","source":"18.(c) Do the token replay conformance checking using your discovered model and the original event log. Does your process model fit?\n","metadata":{"colab_type":"text","id":"Bo5l2xWQXYQ_"}},{"cell_type":"code","source":"token_replay_result = token_based_replay_factory.apply(log, net, initial_marking, final_marking)\nprint(token_replay_result[0])","metadata":{"colab":{},"colab_type":"code","id":"NDuHJDRiXxYQ"},"execution_count":72,"outputs":[{"name":"stdout","output_type":"stream","text":"{'trace_is_fit': True, 'trace_fitness': 1.0, 'activated_transitions': [Assigned Severity, Waiting, Under Processing, Solves, Closed], 'reached_marking': ['sink:1'], 'enabled_transitions_in_marking': set(), 'transitions_with_problems': [], 'missing_tokens': 0, 'consumed_tokens': 5, 'remaining_tokens': 0, 'produced_tokens': 5}\n"}]},{"cell_type":"markdown","source":"Your explanation: \n- Checking the token conformance replay, it is evident that the model fits well with the activated transitions of Assigned Severity->Waiting->Under Processing-> Solves->closed.\n- Trace fits with trace fitness of 1.0 and there are no missing tokens.","metadata":{"colab_type":"text","id":"hf9VsGMEbKAN"}},{"cell_type":"markdown","source":"18.(ci) Calculate the fitness of your model.","metadata":{"colab_type":"text","id":"YzGYxMzGX0rL"}},{"cell_type":"code","source":"# Computing log fitness\nfitness_tokenbasedreplay = replay_fitness_factory.apply(log, net, initial_marking, final_marking)\nprint(fitness_tokenbasedreplay)","metadata":{"colab":{},"colab_type":"code","id":"-Yjjz-IyX6PR"},"execution_count":73,"outputs":[{"name":"stdout","output_type":"stream","text":"{'perc_fit_traces': 82.55312986495782, 'average_trace_fitness': 0.9422531165900576, 'log_fitness': 0.9681466352032891}\n"}]},{"cell_type":"markdown","source":"18.(cii) Are there any deviations between the process model and the event log?","metadata":{"colab_type":"text","id":"wmWIrp8GX_A_"}},{"cell_type":"markdown","source":"Your explanation:\n- There is no significant deviations between the process model and event long as evident from log fitness of 96%","metadata":{"colab_type":"text","id":"Ty8WFiocbFi3"}},{"cell_type":"markdown","source":"### Question 19 (Process Discovery and Conformance Checking using second filtered event log)\n19.(a) Now use the original event log and remove the two most frequent traces, and discover the model based on your new event log(the filtered log without two most frequent traces).\n","metadata":{"colab_type":"text","id":"0PwxekClSCm4"}},{"cell_type":"code","source":"filtered_log2 = variants_filter.apply(log, [max,max1], parameters={\"positive\": False})\nprint(len(filtered_log2))\n\n# Mining for a process tree\n\ntree = inductive_miner.apply_tree(filtered_log2)\nnet, initial_marking, final_marking = inductive_miner.apply(filtered_log2)","metadata":{"colab":{},"colab_type":"code","id":"vPi8IokfSKPS"},"execution_count":74,"outputs":[{"name":"stdout","output_type":"stream","text":"4819\n"}]},{"cell_type":"code","source":"# Petri net visualization\nfrom pm4py.visualization.petrinet import factory as pn_vis_factory\n\ngviz_pn = pn_vis_factory.apply(net, initial_marking, final_marking)\npn_vis_factory.view(gviz_pn)","metadata":{},"execution_count":75,"outputs":[{"execution_count":75,"output_type":"execute_result","data":{"image/png":"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\n","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}]},{"cell_type":"markdown","source":"19.(b) Do the token replay conformance checking using your newly discovered model and the original event log. Does your process model fit?","metadata":{"colab_type":"text","id":"_YKO3A-8SM7A"}},{"cell_type":"code","source":"token_replay_result = token_based_replay_factory.apply(log, net, initial_marking, final_marking)\nprint(token_replay_result[0])","metadata":{"colab":{},"colab_type":"code","id":"IyWVfchYSRUh"},"execution_count":64,"outputs":[{"name":"stdout","output_type":"stream","text":"{'trace_is_fit': False, 'trace_fitness': 0.6400000000000001, 'activated_transitions': [Assigned Severity, Waiting, Under Processing, Solves, Closed], 'reached_marking': ['p_5:1', 'sink:1'], 'enabled_transitions_in_marking': {Waiting, Under Processing}, 'transitions_with_problems': [Waiting], 'missing_tokens': 1, 'consumed_tokens': 5, 'remaining_tokens': 1, 'produced_tokens': 5}\n"}]},{"cell_type":"markdown","source":"Your explanation:\n- Checking the token conformance replay, it is evident that the model does not fits, with the activated transitions of Assigned Severity->Waiting->Under Processing-> Solves->closed. There is 1 missing tokens. The deviation has occurred on transition Waiting.\n- trace fitness is 0.64 ","metadata":{"colab_type":"text","id":"u7y8iFRaa9uJ"}},{"cell_type":"markdown","source":"19.(bi) Calculate the fitness of your model?","metadata":{"colab_type":"text","id":"05uznBXGSVu5"}},{"cell_type":"code","source":"fitness_tokenbasedreplay = replay_fitness_factory.apply(log, net, initial_marking, final_marking)\nprint(fitness_tokenbasedreplay)","metadata":{"colab":{},"colab_type":"code","id":"vmSpVfhoSan3"},"execution_count":65,"outputs":[{"name":"stdout","output_type":"stream","text":"{'perc_fit_traces': 17.44687013504218, 'average_trace_fitness': 0.7237302535510584, 'log_fitness': 0.8492798455914916}\n"}]},{"cell_type":"markdown","source":"19.(bii)  Is there any deviation inside the process model?","metadata":{"colab_type":"text","id":"40v9719GYP8S"}},{"cell_type":"markdown","source":"Your explanation:\n- There is a significant deviations between the process model and event long as evident from log fitness of 84%. There is a deviation that happened at the transition Waiting.","metadata":{"colab_type":"text","id":"secO-k-jayvG"}},{"cell_type":"markdown","source":"### Question 20 (Process Discovery using complete log)\n20.(a) Use the complete event log (original event log) and discover your process model using inductive miner.\n","metadata":{"colab_type":"text","id":"qXDGvC3RSpZu"}},{"cell_type":"code","source":"tree = inductive_miner.apply_tree(log)\nnet, initial_marking, final_marking = inductive_miner.apply(log)","metadata":{"colab":{},"colab_type":"code","id":"eYrEUlz_SmE3"},"execution_count":76,"outputs":[]},{"cell_type":"code","source":"# Petri net visualization\nfrom pm4py.visualization.petrinet import factory as pn_vis_factory\n\ngviz_pn = pn_vis_factory.apply(net, initial_marking, final_marking)\npn_vis_factory.view(gviz_pn)","metadata":{},"execution_count":77,"outputs":[{"execution_count":77,"output_type":"execute_result","data":{"image/png":"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\n","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}]},{"cell_type":"markdown","source":"20.(b) Do the token replay conformance checking using your newly discovered model and the original event log. Does your process model fit?","metadata":{"colab_type":"text","id":"XahMrZNvY0kn"}},{"cell_type":"code","source":"token_replay_result = token_based_replay_factory.apply(log, net, initial_marking, final_marking)\nprint(token_replay_result[0])\n\nfitness_tokenbasedreplay = replay_fitness_factory.apply(log, net, initial_marking, final_marking)\nprint(fitness_tokenbasedreplay)","metadata":{"colab":{},"colab_type":"code","id":"eN-EiMr6Y059"},"execution_count":78,"outputs":[{"name":"stdout","output_type":"stream","text":"{'trace_is_fit': True, 'trace_fitness': 1.0, 'activated_transitions': [Assigned Severity, tauSplit_1, Waiting, Under Processing, tauJoin_2, Solves, Closed], 'reached_marking': ['sink:1'], 'enabled_transitions_in_marking': set(), 'transitions_with_problems': [], 'missing_tokens': 0, 'consumed_tokens': 5, 'remaining_tokens': 0, 'produced_tokens': 5}\n{'perc_fit_traces': 100.0, 'average_trace_fitness': 1.0, 'log_fitness': 1.0}\n"}]},{"cell_type":"markdown","source":"Your explanation: \n- Checking the token conformance replay, it is evident that the model fits, with the activated transitions of Assigned Severity->tau split1->Waiting->Under Processing-> tau split2-> Solves->closed. There are no missing tokens. \n- trace fitness is 1.0\n- log fitness is also 1.0 ","metadata":{"colab_type":"text","id":"b-g6ywFDatOY"}},{"cell_type":"markdown","source":"20.(c) How are these three discovered process models different from each other? Which model is the best fitting to the original log? Why?","metadata":{"colab_type":"text","id":"mqguSN3XY1Px"}},{"cell_type":"markdown","source":"Your explanation:\n- The first process model is without two least frequent traces. The model fits with a fitness of 96%. There is no deviation as well. The second process model is without two most frequent traces. The model does not fit and has a fitness of 84%. There is a deviation and a missing token as well. The third model is the original event log. It fits with a fitness of 100%.\n- Best fitting will be the third model with a fitness of 100%. But considering that the same event log is used to train the model, it is obvious to have a high fitness score. Therefore may also be overfitting. So I think the model 1 with two least frequent traces removed is the better among model 1 and model 2.","metadata":{"colab_type":"text","id":"oTdiAchtY_vg"}}]}