{"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":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-04-30T20:36:18.113531Z","iopub.execute_input":"2023-04-30T20:36:18.114378Z","iopub.status.idle":"2023-04-30T20:36:18.144623Z","shell.execute_reply.started":"2023-04-30T20:36:18.114335Z","shell.execute_reply":"2023-04-30T20:36:18.143516Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Problem definition \nThis dataset contains various information about book reviews left on Goodreads. The goal of this project is to predict the rating of a book based on the data provided. Although the goal of the competition is to perform text analysis to predict the ratings, for this project the predicting features were expanded to include other datasets. A major issue that arose is most of the data is in text format, even date and time data and id values. Due to that limitation I chose the following data sets as features: number of votes on a review, number of comments on a review, and (after data cleaning) the time it took someone to read the book. These three variables were used as the features for 5 models: Linear Regression, Logistic Regression, Decision Tree, Random Forest, and Neural Networks. The Naive Bayes model used the review_text data as its feature. The review_text data was preprocessed using Bag Of Words tokenization and vectorization in order to get the text in numerical format. The target, or variable we are trying to predict, for all models was the rating. The ratings are 6 separate classes (0, 1, 2, 3, 4, 5) so the models had to be able to perform multi-class classification. \n# Motivation\nI work at a used bookstore part-time and it is really hard to stop myself from taking home way too many books. Something I did in the beginning was look at the Goodreads rating of a book to help me decide whether to take it home. Goodreads ratings are so subjective though because our tastes in books are so subjective. Beyond that there are a multitude of factors that influence the ranking of a book: from how well-known it is, to the popularity of the author, to the genre, to how old it is, etc. I am very curious to find some factors that predict the ratings of books and which (if any) factor(s) can help us to predict a books rating.\n# Ethical Concerns about Project Motivation\nThis project seems mostly ethical. A potential ethical concern could be with user privacy, although the reviews are linked to user_ids rather than name data. Another concern come with the implementation or future use of this type of analysis. Some social media sites have started pushing comments that are negative because it keeps people engaged at the detriment of their mental health.\n# Conclusion\nI tried to run predictions on this data using 6 models: Logistic Regression, Decision Tree, Random Forest, Neural Network, and Multinomial Naive-Bayes. The first four models I tried to do predictions on the numerical data availiable in the dataset, even converting from a textual representation of time to a datatime format. However, the most successful model was the model that ran predictions with text, Naive-Bayes Sentiment analysis. The accuracy score was 0.5005. Although better than the other models it still does not meet the 0.80 accuracy score guideline to be a valid preditive model. Since the final model (Multinomial Naive-Bayes with Sentiment Analysis) ran predictions using the review_text as the feature and the rating as the target I was able to submit it in the competition. Its accuracy was **0.49687**.\n# Future Modeling\nIn the future, the models should be retested using the vectorized review text. Different methods can be tested to process the text into numerical data including Term Frequency -Inverse Document Frequency (Tf-IDF) and Word2Vec (W2V). The Neural Network would be interesting to feed the tokenized text to as a feature. Furthermore, Naive Bayes could be used to classify the text into predefined categories, those category arrays could be fed as features to the previously used models as well to see if there is a link between the review category and the rating. Keras is another model that could be used. It is an advanced neural network.\n# References\nData Visualizing: E. Fouda, (2023). EGC493-01 Lecture 3: Regression [PowerPoint slides]., E. Fouda, (2023). EGC493-01 Lecture 13-14: Tableau [Brightspace].\n\nData Cleaning: E. Fouda, (2023). EGC493-01 Lecture 3: Regression [PowerPoint slides].\n\nLinear Regression: E. Fouda, (2023). EGC493-01 Lecture 3: Regression [PowerPoint slides].\n\nLogistic Regression: E. Fouda, (2023). EGC493-01 Lecture 3: Regression [PowerPoint slides].\n\nDecision Tree: E. Fouda, (2023). EGC493-01 Lecture 6: Decision Tree Model [PowerPoint slides].\n\nRandom Forest: E. Fouda, (2023). EGC493-01 Lecture 8: Random Forest [PowerPoint slides].\n\nNeural Network: E. Fouda, (2023). EGC493-01 Lecture 9: Neural Network [PowerPoint slides].\n\nMultinomal Naive-Bayes for Sentiment Analysis: M. Sharma, “Sentiment Analysis (Introduction to Naive Bayes Algorithm),” Medium, 11-May-2020. [Online]. Available: https://towardsdatascience.com/sentiment-analysis-introduction-to-naive-bayes-algorithm-96831d77ac91. [Accessed: 28-Apr-2023]. ","metadata":{}},{"cell_type":"markdown","source":"# Initial Data Visualization and Cleaning\n\nI used df.head() to visualize the set of data given and used .shape to determine how many unique books and users there are. **There are a total of 900000 rows of unique data.**\n\nI attempted to convert the user_id from text data to numerical data so they could be used in models. The data is string representation of hex numbers. I tried converting directly from string to int, and using one-hot encoding. Converting from string to int threw up an error immediately, one-hot encoding ran for a while, then crashed Kaggle becuase it tried to use too much memory.\n\nI decided to clean some data in order to use the amount of time it took the user to read to book to help predict the rating. In order to do that I first had to convert from string format to a datetime format. This required knowledge of the format of the dates, seen in the df.head() printout, and the codes for format. For example, %a represents an abbreviated day of the week, like Sun, Mon, Tues etc.. The times were timezone-aware and had to be converted to utc in order to perform the subtraction. I was able to perform subtraction and place the result in a new column. The null rows were removed so that models could use the data.\n\nI did basic chart visualization of data in a scatter matrix and using Tableau. I also used Tableau to focus on the relationship between the feature data and the target data. You can see that the trend for everage number of votes and average number of comments compared to the rating follows a similar trend. This similarity of data unfortunately does not allow us to gain a broader view of the dataset. If these features are good predictors it doesn't matter as much but if they are mediocre predictors we will not gain all that much more predictation accuracy by including both. You can also see that except for reviews with rating = 1, the number of votes and comments is similar for all ratings. You can see that there is not a huge correlation between the data selected as features, therefore it is expected that the models will not have very high accuracy for their predictions.\n\n![image.png](attachment:dcf2918c-df75-4e1a-adcf-b3bb04d06189.png)\n![image.png](attachment:716ddea4-8293-4691-8947-f291ed457fed.png)\n![image.png](attachment:392118ac-3723-485d-abe9-0b37adc4064e.png)\n![image.png](attachment:10af98de-af4b-4aa9-a880-b7407ff8f377.png)\n","metadata":{},"attachments":{"392118ac-3723-485d-abe9-0b37adc4064e.png":{"image/png":"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"},"10af98de-af4b-4aa9-a880-b7407ff8f377.png":{"image/png":"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"},"dcf2918c-df75-4e1a-adcf-b3bb04d06189.png":{"image/png":"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"},"716ddea4-8293-4691-8947-f291ed457fed.png":{"image/png":"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"}}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nimport struct\nimport matplotlib.pyplot as plt\n\ndf = pd.read_csv('/kaggle/input/goodreads-books-reviews-290312/goodreads_train.csv')\nprint(df.head())\n\n#determine how many unique books\nprint(\"Shape of book_id\", df['book_id'].shape) \nprint(\"No.of.unique book_id values :\", pd.unique(df['book_id']).shape)\n#determine how many unique reviewers\nprint(\"Shape of user_id\", df['user_id'].shape) \nprint(\"No.of.unique user_id values :\", pd.unique(df['user_id']).shape)\n\n#Unsuccessful attempts to convert the id strings to ints so they could be used in testing. The closest method was one-hot encoding. However that used too much memeory and crashed Kaggle.\n'''\n#Create df : dropped hex strings, instead have the int IDs\ndf = df_og.drop('review_id', axis=1) \ndf = df_og.drop('user_id', axis=1) \n\n#Clean user_ID and reviewID data -> convert from hex strings to ints\n#create dummie column of user IDs that are numerical\nuser_idN = pd.get_dummies(df_og.user_id) # get the dummies and store it in a variable\ndf = pd.concat([df, user_idN], axis='columns') # Concatenate the dummies to original dataframe\n\n#create dummie column of review IDs that are numerical\nreview_idN = pd.get_dummies(df_og.review_id) # get the dummies and store it in a variable\ndf = pd.concat([df, review_idN], axis='columns') # Concatenate the dummies to original dataframe\nprint(df.head())\n'''\n\n#Data cleaning to obtain a new column with time data as numeric\ndf = df.dropna() #remove the null rows before we try converting\n#convert from string date and time to a datetime object and convert times to UTC so we can perform subtraction\ndf['dt_read_at'] = pd.to_datetime(df['read_at'], format='%a %b %d %H:%M:%S %z %Y', utc = True)\ndf['dt_started_at'] = pd.to_datetime(df['started_at'], format='%a %b %d %H:%M:%S %z %Y', errors = 'coerce', utc=True) \n\ndf = df.dropna() #remove the null rows produced by converting \n\n#created a new numerical column with the amount of time it took to read the book as a float\ndf['timeToRead'] = df['dt_read_at'] - df['dt_started_at']\ndf['timeToRead'] = df['timeToRead'] / np.timedelta64(1, 'D')\n\nprint(df.head()) #show df with our new column to check output\n\n\n#Visualize Data\npd.plotting.scatter_matrix(df,0.5,(20,25)) #basic data visualization\nplt.show() ","metadata":{"_kg_hide-input":false,"_kg_hide-output":false,"execution":{"iopub.status.busy":"2023-04-30T20:36:18.146560Z","iopub.execute_input":"2023-04-30T20:36:18.146934Z","iopub.status.idle":"2023-04-30T20:38:02.708659Z","shell.execute_reply.started":"2023-04-30T20:36:18.146902Z","shell.execute_reply":"2023-04-30T20:38:02.707651Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Clean and Prep Data for Models\n\n**Data Cleaning: Create our timeToRead series**\n\nRather than taking the time, power, and memory to create the timeToRead column each time, this code can be run once before running the other models.\n\n**Assign the features targets and split the data**\n\nAssigns the features as n_comments, n_votes, and timeToRead and the target as the rating. The split is 80% training data, 20% test data. random_state=42 was assigned so the split would be the same on every run.\n\nThe models that do not use this dataset/data split can rewrite within their code. THeir variables were named with reference to their model so they would not overwrite our \"global\" X_train, X_test, y_train, y_test variables on full notebook runs.","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nfrom sklearn.model_selection import train_test_split \n\ndf = pd.read_csv('/kaggle/input/goodreads-books-reviews-290312/goodreads_train.csv')\n\n#convert from string date and time to a datetime object and convert times to UTC so we can perform subtraction\ndf['dt_read_at'] = pd.to_datetime(df['read_at'], format='%a %b %d %H:%M:%S %z %Y', utc = True)\ndf['dt_started_at'] = pd.to_datetime(df['started_at'], format='%a %b %d %H:%M:%S %z %Y', errors = 'coerce', utc=True) \n\ndf = df.dropna() #remove the null rows produced by converting \n\n#created a new numerical column with the amount of time it took to read the book as a float\ndf['timeToRead'] = df['dt_read_at'] - df['dt_started_at']\ndf['timeToRead'] = df['timeToRead'] / np.timedelta64(1, 'D')\n\n#assign features and target\nfeature_names = ['n_votes', 'n_comments', 'timeToRead']\nX = df[feature_names].values\ny = df['rating'].values\n\n# Split the data into training and testing sets\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)","metadata":{"execution":{"iopub.status.busy":"2023-04-30T20:38:02.710938Z","iopub.execute_input":"2023-04-30T20:38:02.711360Z","iopub.status.idle":"2023-04-30T20:38:59.631689Z","shell.execute_reply.started":"2023-04-30T20:38:02.711325Z","shell.execute_reply":"2023-04-30T20:38:59.630383Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Linear Regression Model:\n\nLinear Regression is used for regression problems and Logistic is used for Classification. \nBecause the ratings are broken into discrete values (0, 1, 2, 3, 4, 5), this is a multiclass classification problem. This model should perform worse than our multiclass-classification models.\n\n\nLinear regression model requires all numeric values.\nx vars (features): n_votes, n_comments, timeToRead\ny var (target): rating\nCreated three LR models with various combinations of x variables:\nn_votes and n_comments vs. rating,\nn_votes vs. rating,\nn_comments vs. rating, and\nn_votes, n_comments, and timeToRead vs. rating\n\nI chose to run these tests because I was curious if adding an extra frame but reducing the amount of data (because we removed more rows with null values) produced better results. \n\nData Cleaning: For the models built with only one feature, the array had to be reshaped in order to run.\n\n\nModel Evaluation: \nModel Evaluation was performed using the r2 score to determine accuracy. R2 score is used for scoring regression, accuracy is used for scoring classification problems. R2 score ranges from 0-1 with 1 being perfect. R2 score could be negative if the prediction was worse than random.\n\nAll of the models are bad. **The R2 score shows us that they are barely better than random chance**. **The best option was the one with the most data, where the x values are number of votes, number of comments, and the amount of time it took to read the book with an r2 score of 0.0006...**. The worst was using just number of votes. However from this we can tell that **the time it took to read the book and the number of votes and the number of comments on a review are not good predictors of the rating in the Linear Regression model.**\n\nA confusion matrix does not make sense for model evaulation since we do not have only two states and the True Positive, False Positive, True Negative, False Negative doesnt really matter for this type of problem.","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nimport struct\nimport matplotlib.pyplot as plt\nfrom sklearn.linear_model import LinearRegression #import Linear Progression model\nfrom sklearn.model_selection import train_test_split \n# importing r2_score module\nfrom sklearn.metrics import r2_score\n\ndf2 = pd.read_csv('/kaggle/input/goodreads-books-reviews-290312/goodreads_train.csv')\n\n#Linear regression model comparing the three numerical datasets\nX1 = df2[['n_votes', 'n_comments']].values\nyLR = df2['rating'].values\n\n# Split the data into training and testing sets\nX1_train, X1_test, y_trainLR, y_testLR = train_test_split(X1, yLR, test_size=0.2, random_state=42)\n\n# create model and fitting the training data\nLR = LinearRegression()\nLR.fit(X1_train,y_trainLR)\n\n#predition\ny1_prediction = LR.predict(X1_test)\nprint(\"y1_prediction- x values nVotes and nComments\", y1_prediction)\n\n###################################################################################################\n\n#Linear regression model for n_votes to rating\nX2 = df2['n_votes'].values\nX2 = X2.reshape(-1, 1)\n\n# Split the data into training and testing sets\nX2_train, X2_test, y_trainLR, y_testLR = train_test_split(X2, yLR, test_size=0.2, random_state=42)\n\n# create model and fitting the training data\nLR.fit(X2_train,y_trainLR)\n\n#predition\ny2_prediction = LR.predict(X2_test)\nprint(\"y2_prediction- x values nVotes\", y2_prediction)\n\n###################################################################################################\n\n#Linear regression model for n_comments to rating\nX3 = df2['n_comments'].values\nX3 = X3.reshape(-1, 1)\n\n# Split the data into training and testing sets\nX3_train, X3_test, y_trainLR, y_testLR = train_test_split(X3, yLR, test_size=0.2, random_state=42)\n\n# create model and fitting the training data\nLR.fit(X3_train,y_trainLR)\n\n#predition\ny3_prediction = LR.predict(X3_test)\nprint(\"y3_prediction- x values nComments\", y3_prediction)\n\n#####################################################################################################\n\n#Linear regression model for n_comments, n_votes, and timeToRead to rating\nX4 = df[['n_votes', 'n_comments', 'timeToRead']].values\ny4 = df['rating'].values\n\n# Split the data into training and testing sets\nX4_train, X4_test, y4_train, y4_test = train_test_split(X4, y4, test_size=0.2, random_state=42)\n\n# create model and fitting the training data\nLR.fit(X4_train,y4_train)\n\n#predition\ny4_prediction = LR.predict(X4_test)\nprint(\"y4_prediction- x values nVotes nComments timeToRead\", y4_prediction)\n\n#####################################################################################################\n#####################################################################################################\n\n\n#check accuracy\nscore1=r2_score(y_testLR,y1_prediction)\nprint('nVotes and nComments: r2 score is ',score1)\n\nscore2=r2_score(y_testLR,y2_prediction)\nprint('nVotes: r2 score is ',score2)\n\nscore3=r2_score(y_testLR,y3_prediction)\nprint('nComments: r2 score is ',score3)\n\nscore4=r2_score(y4_test,y4_prediction)\nprint('nComments nVotes timeToRead: r2 score is ',score4)","metadata":{"execution":{"iopub.status.busy":"2023-04-30T20:38:59.633990Z","iopub.execute_input":"2023-04-30T20:38:59.634495Z","iopub.status.idle":"2023-04-30T20:39:12.336421Z","shell.execute_reply.started":"2023-04-30T20:38:59.634446Z","shell.execute_reply":"2023-04-30T20:39:12.334988Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Logistic Regression\n\nLogisitic Regression is similar to Linear Regression but built for classification problems. \n\nInitially, an error occured because the Logistic Regression tried to iterate more time than allowed. To handle this error, the LogisticRegression model was given a new max_iter value of 1000. This allowed the model to run.\n\nBy viewing the prediction array from the previous Linear Regression model with the same features, you can see that the Logistic Regression model essentially consists of the rounded values from the Linear Regression Model. When you use r2 to score this model, the result is -48.17834004992871, much worse than random chance. This shows us how important it is to use the proper accuacy scoring method for each model.\n \n \n**The accuracy is 0.36429731925264014. This is much higher than the accuracy of the Linear regression model. That makes sense since this is a classification problem.** The precision measures accuracy of true positives. For this model the precision is 0.16877625367925644. The recall measures the completeness of predicted positives. The recall was 0.12920373607207344.","metadata":{}},{"cell_type":"code","source":"from sklearn.linear_model import LogisticRegression\nfrom sklearn.metrics import r2_score, recall_score, precision_score\n\n# create model and fitting the training data\nLR = LogisticRegression(max_iter=1000)\nLR.fit(X_train,y_train)\n\n#predition\nprediction = LR.predict(X_test)\nprint(prediction)\n\n#check accuracy\nprint('accuracy:',LR.score(X_test, y_test))\n#check precision\nprint('precision:', precision_score(prediction, y_test, average = 'macro'))\n#check recall\nprint('recall:',recall_score(prediction, y_test, average = 'macro', zero_division = 0))\n#r2 score\nprint('r2 score: ', r2_score(prediction, y_test))","metadata":{"execution":{"iopub.status.busy":"2023-04-30T20:39:12.344349Z","iopub.execute_input":"2023-04-30T20:39:12.348010Z","iopub.status.idle":"2023-04-30T20:42:26.878729Z","shell.execute_reply.started":"2023-04-30T20:39:12.347940Z","shell.execute_reply":"2023-04-30T20:42:26.876454Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Decision Tree Model:\nA decision tree model is helpful for visualization. In this case I used a multiclassification DT model. The features used were the most successful model for Linear Regression, n_comments, n_votes, and timeToRead. Offering more features is better for initial testing. Each rating is discrete so it could be considered as a class. One issue that arose when fitting to this model was a value error. It defaults to finding the binary average which is not possible with a multiclass model. I chose the weighted average because it takes the average for the classes (like macro) then weights it to account for a class imbalance (like micro).\n\nOne problem with Decision Trees is that they are prone to overfitting. To counteract that I ran a gridsearch to determine a good set of parameters to restrict our decision tree with. On initial model-building I only used the features n_votes and n_comments and recieved best params of {'max_depth': 3, 'max_leaf_nodes': 10, 'min_samples_leaf': 6} \nHowever, on rebuilding with the additional feature of timeToRead I got best_params= {'max_depth': 6, 'max_leaf_nodes': 10, 'min_samples_leaf': 1}\n\nThe test-train split was done using KFolds. I calculated accuracy, precision, and recall for a tree built with gini, and for a tree built with entropy and the result was nearly equivalent. You can't use weighted for these features so instead I chose average = 'macro' because when I triedto use average = 'micro' to account for the class imbalance the results for accuracy, precision, and recall were all the same. For my original testing with only 2 features, with best params used the gini and entropy accuracy was about 0.355. For the decision tree model with all 3 features, with best params used, the gini and entropy **accuracy was about 0.369**. **Precision and recall scores approximately 0.15.** **The Decision Tree outperformed the Linear Regression model and performed about as well as the Logistic Regression model. However, it is still not a good enough model to use for predictions.**\n\nI printed out the decision tree for visualization. Because the gini scores were often slightly higher, I switched the order of the loop to test gini last so the final DT is the gini model which will get printed. ","metadata":{}},{"cell_type":"code","source":"from sklearn.tree import DecisionTreeClassifier\nfrom sklearn.model_selection import KFold\nfrom sklearn.metrics import accuracy_score, precision_score, recall_score, make_scorer, f1_score\nfrom sklearn.model_selection import GridSearchCV\n\n#determine best params\nparam_grid = {\n'max_depth': [3, 4, 6, 8, 10],\n'min_samples_leaf': [1, 3, 6],\n'max_leaf_nodes': [5, 10, 15]}\ndt0 = DecisionTreeClassifier()\nscorer = make_scorer(f1_score, average = 'weighted')\ngs = GridSearchCV(dt0, param_grid, scoring=scorer, cv=5)\ngs.fit(X, y)\nprint(\"best params:\", gs.best_params_)\n\n#split tree into train/test\n#analyze tree as gini and entropy with best params\nkf = KFold(n_splits=5, shuffle=True)\nfor criterion in ['entropy', 'gini']:\n    print(\"Decision Tree - {}\".format(criterion))\n    accuracy = []\n    precision = []\n    recall = []\n    for train_index, test_index in kf.split(X):\n        X_trainDT, X_testDT = X[train_index], X[test_index]\n        y_trainDT, y_testDT = y[train_index], y[test_index]\n        dt = DecisionTreeClassifier(criterion=criterion, max_depth=6,max_leaf_nodes=10, min_samples_leaf=1)\n        dt.fit(X_trainDT, y_trainDT)\n        y_pred = dt.predict(X_testDT)\n        accuracy.append(accuracy_score(y_testDT, y_pred))\n        precision.append(precision_score(y_testDT, y_pred, average = 'macro', zero_division = 0))\n        recall.append(recall_score(y_testDT, y_pred, average = 'macro', zero_division = 0))\n    print(\"accuracy:\", np.mean(accuracy))\n    print(\"precision:\", np.mean(precision))\n    print(\"recall:\", np.mean(recall), '\\n')\n    print()\n\n#print tree\nfrom sklearn.tree import export_graphviz\nimport graphviz\nfrom IPython.display import Image\n\ndot_file = export_graphviz(dt, out_file='tree_limited.dot',feature_names=feature_names)\ngraph = graphviz.Source(dot_file)\n\n#export image\n!dot -Tpng tree_limited.dot -o tree_limited.png -Gdpi=600\nImage(filename = 'tree_limited.png')\n","metadata":{"execution":{"iopub.status.busy":"2023-04-30T20:42:26.880009Z","iopub.execute_input":"2023-04-30T20:42:26.880360Z","iopub.status.idle":"2023-04-30T20:44:58.555755Z","shell.execute_reply.started":"2023-04-30T20:42:26.880329Z","shell.execute_reply":"2023-04-30T20:44:58.553844Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Random Forest\n\nA Random Forest model tends to produce good results for any model \"out of the box\" and requires little tuning. I decided to use the same features for this model as I did for the previous two in order to more directly compare the results (n_votes, n_comments, and timeToRead). \n\nSince Random Forests are not prone to overfitting, little tuning is necessary. However, I decided to tune the number of estimators (decision trees). I attempted to tune the n_estimators variable in two different ways, first I ran a grid search and provided 6 discrete numbers to choose from. Because I am using a multiclass dataset with is unbalanced I chose to use weight average f1 scoring. The best n_estimator parameter was dependent on the random_state variable I chose. When random_state = 42 the best n_estimator was 25. Next, I attempted to run a gridsearch and print the elbow graph. This would help determine where adding trees no longer adds much differentialbility to the graph, resulting in an imbalanced processing power to reward relation. However after allowing the code to run for over 30 minutes with no result so I decided to go with the predetermined values in the gridsearch. \n\nSince I am only feeding 3 features I chose not to find the best features.\n\n**The accuracy of the model when using the best param was 0.329**. **The Random Forest model outperformed the Linear Regression model. It performed slightly worse than the Logistic Regression and Decision Tree Models.** This was unexpected as the Random Forest model is usually expected to perform better than the Decision Tree.\n\nThe precision measures accuracy of true positives. For this model the precision is 0.17318395084858437. The recall measures the completeness of predicted positives. The recall was 0.17527302363864591.\n","metadata":{}},{"cell_type":"code","source":"from sklearn.metrics import accuracy_score, make_scorer, f1_score, precision_score, recall_score\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.model_selection import GridSearchCV\n\n#tuning- find best n_estimators within limited set of numbers\nparam_grid = {\n'n_estimators': [10, 20, 25, 30, 35],\n}\nrf = RandomForestClassifier(random_state=101)\nscorer = make_scorer(f1_score, average = 'weighted')\ngs = GridSearchCV(rf, param_grid, scoring=scorer, cv=5)\ngs.fit(X, y)\nprint(\"best params:\", gs.best_params_)\n\n'''#tuning best n_estimated with an elbow graph\nn_estimators = list(range(1, 31))\nparam_grid = {\n    'n_estimators': n_estimators,\n}\nrf = RandomForestClassifier()\ngs = GridSearchCV(rf, param_grid, cv=5)\ngs.fit(X, y)\nprint(\"best params:\", gs.best_params_)\nscores = gs.cv_results_['mean_test_score']\n# [0.91564148, 0.90685413, ...]\n\n#plot our elbow graph to find the best n_estimator\nimport matplotlib.pyplot as plt\n\nscores = gs.cv_results_['mean_test_score']\nplt.plot(n_estimators, scores)\nplt.xlabel(\"n_estimators\")\nplt.ylabel(\"accuracy\")\nplt.xlim(0, 30)\nplt.ylim(0.9, 1)\nplt.show()'''\n\n\n#run RF model  with best n_estimator param of 25 and determine accuracy\nrf = RandomForestClassifier(n_estimators = 25)\nrf.fit(X_train, y_train)\nfirst_row = X_test[0]\nprint(rf.predict(X_test))\nprint(\"random forest accuracy:\", rf.score(X_test, y_test))\n#check precision\nprint('precision:', precision_score(rf.predict(X_test), y_test, average = 'macro'))\n#check recall\nprint('recall:',recall_score(rf.predict(X_test), y_test, average = 'macro'))\n","metadata":{"execution":{"iopub.status.busy":"2023-04-30T20:44:58.557948Z","iopub.execute_input":"2023-04-30T20:44:58.558424Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Neural Network\n\nNeura Networks perform well in cases where we have a lot of features because they automatically do feature engineering without requiring domain knowledge to restructure the features. Neural networks also require numerical data rather than categorical data. Therefore I stuck to using the three features I used for the previous models, n_comments, n_votes, and timeToRead although neural networks perform better on more features. A neural network is especially good for a multiclass problem such as this dataset. In the future, using the tokenized review_text data as a feature for this model would probably be a good idea becaue the tokenization created a huge amount of features (approximately 200,000).\n\nThe default mlp performed almost exactly as well as the tweaked mlp but significantly increased the power and runtime. The **mlp accuracy score** on the tweaked version was **0.369910641754671**. The mlp accuracy score on the untweaked version was 0.369553208773355. The default performed better than tweaked version and required less time.\n\nThis model performed about as well as the Logistic Regression, Decision Tree, and Random Forest Models did with the features provided. It seems that **0.37 is approximately the max prediction we can gain using n_votes, n_comments, and timeToRead.** These **features seem to be nearly irrelavant to predicting the rating** left by the user on the book.","metadata":{}},{"cell_type":"code","source":"from sklearn.neural_network import MLPClassifier\nfrom sklearn.metrics import precision_score, recall_score\n\nmlp = MLPClassifier()\nmlp.fit(X_train, y_train)\n\n#print prediction array\nprint(\"untweaked prediction array:\",mlp.predict(X_test))\n#score accuracy\nprint(\"untweaked accuracy:\", mlp.score(X_test, y_test))\n#check precision\nprint('untweaked precision:', precision_score(mlp.predict(X_test), y_test, average = 'macro'))\n#check recall\nprint('untweaked recall:',recall_score(mlp.predict(X_test), y_test, average = 'macro', zero_division = 0))\n\n#test tweaked mlp values to find best accuracy with varied hidden layer sizes\nmlp = MLPClassifier(hidden_layer_sizes=(100, 50, 15), alpha=0.0001, solver='adam', random_state=42)\nmlp.fit(X_train, y_train)\n\n#print tweaked prediction array\nprint(\"tweaked prediction array:\",mlp.predict(X_test))\n#tweaked accuracy score\nprint(\"tweaked accuracy:\", mlp.score(X_test, y_test))\n#check precision\nprint('tweaked precision:', precision_score(mlp.predict(X_test), y_test, average = 'macro'))\n#check recall\nprint('tweaked recall:',recall_score(mlp.predict(X_test), y_test, average = 'macro', zero_division = 0))","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Naive Bayes Sentiment Analysis\n\nNaive-Bayes is a good model for text analysis. It works by looking at the occurance of a term and the frequency of its occurance in relation to the overall text size. Naive Bayes is better at text classfication than sentiment analysis. Our predictions are on \"sentiment\" analysis i.e. ratings from 0-5 stars so sentiment analysis makes more sense.\n\nIn order to use Naive-Bayes for sentiment analysis the text was classified using the bag of words (BOW) model. BOW converts text to numbers that can be used in an algorithm (Naive-Bayes). BOW returns a vector with the word and the number of times the word has been used before. You can use different n_grams for vectorizing, meaning the model with take in n_words at a time. Uses stopwords = 'english' to remove erronious common words such as 'i' 'me' 'my' 'myself' 'we' etc. so it doesn't take too much memory.\n\nModel Tweaking: **The accuracy for the monogram vectorization Naive-Bayes model was 0.50005**. By changing the ngram_range to (2,2) the model was retested as a a bigram vectorized model. The accuracy for the bigram vectorization model was 0.4852611111111111. The trigram vectorization took way longer and crashed kaggle because it used too much memory.\n\n**While this accuracy was not good enough to use for prediction, it is the best accuracy of any model.**\n\n\nThis model was used to make a prediction for the real test set. Originally an error message occured when trying to run a prediction on the new data. *ValueError: X has 216238 features, but MultinomialNB is expecting 291357 features as input.* This error occured because the vectorization of text created an inconsistant number of features for the training set and testing set. This error was handled by using cv.transform on the test data rather than cv.fit_transform to tokenize the text. The accuracy when running against the test dataset was 0.49687.\n\n\nUsed this website as a guide.\nhttps://towardsdatascience.com/sentiment-analysis-introduction-to-naive-bayes-algorithm-96831d77ac91","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nfrom sklearn.feature_extraction.text import CountVectorizer\nfrom nltk.tokenize import RegexpTokenizer\nfrom sklearn.naive_bayes import MultinomialNB\nfrom sklearn.model_selection import train_test_split\n\ndf_NB = pd.read_csv('/kaggle/input/goodreads-books-reviews-290312/goodreads_train.csv')\n\n#tokenize text to prep for parsing Bag of Words model\ntoken = RegexpTokenizer(r'[a-zA-Z0-9]+') #sysmbols for tokenizing\ncv = CountVectorizer(stop_words='english',ngram_range = (1,1),tokenizer = token.tokenize) #define count vectorizer\ntext_counts = cv.fit_transform(df_NB['review_text'])\n\n#Prep training set data for model\n#split test/train\nX_trainNB, X_testNB, Y_train, Y_test = train_test_split(text_counts, df_NB['rating'], test_size = 0.2, random_state = 42)\n\n#Utilizing Multinomial Naive-Bayes model\n#fit the model\nMNB = MultinomialNB()\nMNB.fit(X_trainNB, Y_train)\n\n#evaluate model\npredicted = MNB.predict(X_testNB)  #get array of predictions\nprint(predicted)\nprint(\"MNB accuracy score:\", MNB.score(X_testNB, Y_test))\nprint('precision:', precision_score(predicted, Y_test, average = 'macro'))\nprint('recall:',recall_score(predicted, Y_test, average = 'macro'))\n\n#Test Data against MNB model\n#import testing dataset\ndf_test = pd.read_csv('/kaggle/input/goodreads-books-reviews-290312/goodreads_test.csv')\n\n#tokenize test dataset text \n#must use .transform, rather than fit_transform so the model can match number of features\ntest_text_counts = cv.transform(df_test['review_text'])\n\n#using model predict based on the target test data\nmy_prediction = MNB.predict(test_text_counts)\n\n#view predicted ratings\nprint(\"prediction on real test text\",my_prediction)\n\n#prep data for submission to .csv\n#create 'rating' column filled with my_prediction\ndf_test['rating'] = my_prediction\n#choose id and rating columns to place in submission.csv\ndf_test[['review_id', 'rating']].to_csv('submission.csv', index=False)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}