Gist 5679d9c117e4713bbd72f666306c2ba9

Created on Cognitive Class Labs

This example demonstrates simple linear regression on a dataset of fuel consumption and CO2 emissions, plotting the relationship between engine size and emissions. The code uses scikit-learn to split the data into training and test sets, then fits a regression line to predict emission values. Using matplotlib and pylab, it visualizes the training data, the regression line, and the test data points to evaluate the model's accuracy. The dataset originates from a CSV file downloaded via wget from IBM Object Storage.

DDAVINDER

61% match

Linear Model

The visualization generates random points and a grid of squares, with each square’s fill color encoding the residual error of the best-fit line passing through that square’s center. A line is drawn across the chart; hovering over any square updates this line to show the fitted linear model with the minimum squared error for points near that location. The code uses d3.v3 with SVG, a linear color scale, and mouseover events to dynamically redraw the line.

NNPashaP

61% match

Gist 9963650

This visualization demonstrates how linear calibration transforms MRI brain volume measurements from two scanner sites, comparing simulated models that combine variables, add site regressors, or calibrate data to evaluate equivalence between scanners. It uses Python with numpy, matplotlib, seaborn, statsmodels, scipy, and rpy2 to run ordinary least squares and robust linear models, while rendering plots inline via IPython Notebook.

59% match

Gist c432d79c438c772e0332

This visualization demonstrates a regression analysis workflow from a Julia notebook, showing how a random feature map (phi) transforms 2D samples into a 250-dimensional space for least-squares fitting. The code computes the optimal weight vector w using normal equations with ridge regularization, evaluating the objective function across different lambda values. It uses Julia’s linear algebra operations (`transpose`, `\` for matrix division, `eye`) and `randn` for Gaussian sampling, with SVG rendering and animation to display the fitted model’s error surface and weight trajectories.

Gist 1c87e6026e22589f1729

This visualization tracks the misclassification rate of a stochastic gradient descent classifier over 5,000 iterations, with separate lines for training and test datasets. The code uses Python with NumPy, Pandas, and Seaborn’s `color_palette` for styling, while Matplotlib’s `plot` and `legend` functions render the two curves. The y-axis shows misclassification rate and the x-axis shows iteration counter, revealing how the model’s performance evolves during a single pass through the data with a constant step size.

Gist 9540056

This visualization simulates multi-site MRI data, modeling brain volume across two sites for control and patient groups, and compares three regression strategies for combining site data, including a linear calibration transformation. It shows how model predictions and confidence intervals vary across age and disease metrics. Built in a Jupyter notebook, the code uses NumPy for data generation, statsmodels for ordinary least squares and robust linear model fitting, and matplotlib with prettyplotlib for plotting the results.

Gist d4b2a37f3ec9a90469c8

This visualization simulates longitudinal brain-imaging measurements across multiple subjects and time points, showing how percentile values (e.g., MTR_25) vary and are modeled with ordinary least squares regression per subject. It first computes subject means, average standard deviations, and true beta coefficients from the input CSV, then generates synthetic data for three time points using normal distributions. The code relies on pandas for data manipulation, statsmodels formula API for OLS fitting, numpy for random sampling, and seaborn for plotting.

Gist 130517

This visualization displays a sparse, grid-based dataset from a gist by DylanFM, using a fixed-size coordinate system (values such as 1056, 64, 1152, and 4608) to plot points in a two-dimensional space. The rendering relies on D3’s `d3.forceSimulation` for positioning and likely `d3.geoPath` or a similar path generator to draw shapes, though the specific API is not explicitly shown. The data, provided as a plain text file, is read and mapped directly to screen coordinates without additional transformation, creating a minimal scatter-like plot. The example emphasizes raw data plotting with minimal abstraction, focusing on the direct conversion of numeric inputs into visual marks.

DDylanFM