Gist 392220bb33ff81da26fe5b1787cd505b

Gist f2b16fc0ccd955514d00fbf44683516f

This example demonstrates a temporal split of simulated financial transaction data into training and test sets for fraud detection. The code loads transaction records spanning from July 25 to August 14, 2018, then partitions them into three consecutive seven-day windows: training, delay, and test periods. It uses the custom `read_from_files` function to ingest the data and `get_train_test_set` to perform the split, while counting fraudulent transactions with the `TX_FRAUD` column. The visualization relies on Pandas DataFrames and Python’s `datetime` module to manage the time-based segmentation.

YYannael

74% match

Gist b5bb72e72049a84c7a5a0b3f0e1a3498

The example shows how to load and prepare simulated credit-card transaction data for fraud-detection analysis. It initializes a Python environment by downloading shared helper functions and cloning a GitHub repository containing pre-transformed datasets. The code uses the `%run` magic command to execute `shared_functions.py` and checks for local data before cloning, ensuring reproducible access to the simulated data source. This setup is a prerequisite for subsequent visualization steps, though no chart or rendering API is invoked in this initialization phase.

YYannael

64% match

tut21

This example shows a supervised machine learning workflow using the random forest algorithm to predict income categories from a dataset of demographic features. The code builds a random forest model with 500 trees and three variables considered at each split, using the `randomForest` function from the R randomForest package. After training on a designated training set, it generates predictions and compares them against actual income labels using a confusion matrix, reporting a classification accuracy of approximately 77.6%. The visualization itself is not graphical; instead, it presents the model’s performance metrics as tabular and numeric output.

57% match

tut19

This visualization demonstrates the accuracy of a decision tree model in predicting the `INCOME` category from a training dataset. It shows a confusion matrix and the computed accuracy rate of approximately 73.4%, comparing predicted versus actual income labels. The R code uses the `caret` package’s `train()` function with the `"rpart"` method to build the model, then evaluates it with the `table()` and `mean()` functions. The script also loads `e1071`, though the visible steps primarily rely on caret and rpart for classification and validation.

57% match

Gist 15ec54668ea15cb0eada554bae7d6f13

This example demonstrates the concept of precision at k by generating a synthetic dataset of 1,000 “potential bad actors,” each labeled as fraudulent or not, and then computing precision for the top 40 ranked predictions based on a model’s probability scores. The code uses `faker` to create fake names and emails, `numpy.random` for random labeling, and `scipy.stats.truncnorm` to simulate skewed probabilities. It defines a custom `precision_at_k` function that sorts scores in descending order and calculates the fraction of true positives among the first *k* entries, while also comparing it to scikit-learn’s `average_precision_score` for context.

BBrideau

tut7

This visualization demonstrates the outcome of a classification tree model for predicting Brexit voting behavior. It displays the fitted decision tree structure, showing variable importance and split values. The example uses the R programming language with the `printcp` and `plotcp` functions from the `rpart` package to output the model’s complexity parameters and render its cost-complexity plot. The data originates from a gist by BenHeubl, with the underlying source file containing the model fitting code.

tut8

This visualization shows a decision-tree model of Brexit voting patterns, mapping how various factors predict the percentage of Leave votes across the UK. The code builds a classification tree using the `rpart` function on the `Brexit` dataset, with `Percent.Leave` as the target variable. The tree’s branching structure reveals the key demographic or geographic splits that separate Remain- and Leave-leaning areas. The rendering relies on R’s base graphics to display the tree nodes and splits, while the user must supply a complexity parameter (`cp`) to control tree depth and overfitting.

tut5

This example shows a decision tree model of Brexit-related data, visualizing the hierarchical splitting of predictor variables into classification nodes. The tree uses colored boxes with a red-to-yellow-to-green palette to indicate different outcome probabilities or class assignments across branches. The plot is generated with the `rpart.plot` function, using type 4 for a fancier layout, extra 0 for minimal node labeling, and dashed branch lines via `branch.lty = 3`. The underlying model, `Brexit_model`, is built from an rpart decision tree, and the visualization directly renders its structure without additional data processing.