Skip to main content
100%

Gist 6bdb29f6cdeb680f0306

✓ Published0🌍 Public
Aaman-tiwari
Last edited Mar 15, 2016
Created on Mar 15, 2016

This example demonstrates a method for performing nearest-neighbor upsampling on a NumPy array using matrix multiplications, effectively doubling the spatial dimensions of the data. The visualization shows how the original input values are expanded by inserting repeated rows and columns through carefully constructed kernel matrices. The `unpool` function builds two tile-based kernels, `kernel_x` and `kernel_y`, and applies them via sequential `np.matmul` operations with transposes to correctly broadcast the expansion. This approach leverages linear algebra primitives from NumPy without explicit looping or image-specific libraries.

AI-generated description

Similar vizzes

Loading thumbnail…

Gist 585271

This example demonstrates a one-dimensional Gaussian kernel generated by the Ruby class `GThang`, which computes a discrete approximation of the kernel over a radius of three units. It shows the resulting array of kernel values printed to the console via `puts`, highlighting the symmetric, bell-shaped weighting used in kernel density estimation. The code implements the Gaussian function manually using `Math.sqrt`, `Math::PI`, and `Math::E`, with no external visualization libraries or data sources. The output is a simple numeric list, not a graphical plot, so the visualization is limited to the textual representation of the kernel’s weights.

XXXX
51% match
Loading thumbnail…

Gist 740405b6405a6644fd16d3a50588de51

This example shows a step-by-step transformation of the classic wine dataset, progressively filtering and selecting columns to isolate alcohol content, color intensity, and hue before randomly sampling five rows. The code chains pandas operations—including `rename`, `assign`, `query`, and `sort_values`—through a custom `csnap` helper that snapshots the intermediate state after each step. It uses NumPy’s `np.where` to create a binary `color_filter` column based on hue and color intensity thresholds, then resets the index and samples five rows with `sample(5)`. The visualization highlights how pipelined data wrangling can be inspected at each stage.

Aadiamaan92
50% match
Loading thumbnail…

Gist b2dba9d3a950202a27aaf70eb500b55b

This example visualizes the relationships between multiple MRI scan sessions by computing and plotting the affine and warp transformations needed to align them to a common template. The notebook uses Python with pandas, numpy, and matplotlib, along with custom functions built on nipype’s file management utilities, to load alignment status JSONs and construct transformation dictionaries. It shows how each session’s T1-weighted image can be mapped to any other session through a chain of affine and warp operations, with the resulting paths displayed as a network of connections. The rendering relies on matplotlib’s inline plotting to present the computed transformation graph.

Aakeshavan
50% match
Loading thumbnail…

Gist 0383cb7786315e4c8a11

This example demonstrates fitting a Gaussian mixture model to multidimensional data using the expectation-maximization (EM) algorithm. It shows how the model estimates means, standard deviations, and mixture proportions for each dimension of the dataset. The code uses pandas to read the CSV file, NumPy for numerical operations, and matplotlib for inline plotting. It applies the EM algorithm separately to each column of the data, then prints the fitted parameters for each dimension.

Aakeshavan
50% match
Loading thumbnail…

Perform sliding window analysis on count data in a 2d array

This example visualizes the spatial distribution of spot counts and total intensity in a 2D array using sliding-window binning. It shows two heatmaps side-by-side, one for spot count and one for total intensity, with the y-axis inverted to match image orientation, plus a joint regression plot revealing the correlation between the two metrics. The code uses pandas for data loading and quantile-based thresholding, then bins the filtered data into 10x10 pixel windows. Seaborn’s `heatmap` and `jointplot` render the visualizations, while `matplotlib.pyplot` handles the subplot layout and axis inversion.

FFloWuenne
50% match
Loading thumbnail…

configure random transformations and normalization operations to be done on your image data during training

The example demonstrates how Keras’s `ImageDataGenerator` applies random geometric transformations and normalization to a single input image, producing a grid of augmented variations saved as JPEG files. It shows the effects of rotation, width and height shifts, shear, zoom, and horizontal flipping on a cat photo, with pixel fill mode set to nearest. The code loads an image via `load_img`, converts it to a NumPy array with `img_to_array`, reshapes it into a batch, and then uses the generator’s `.flow()` method to generate and save 20 preview images to a directory. The approach relies entirely on Keras’s preprocessing API, with no custom data source or rendering.

Aabehmiel
50% match
Loading thumbnail…

Gist 5679d9c117e4713bbd72f666306c2ba9

This example demonstrates linear regression on a small dataset, plotting training and testing points to show the relationship between the input and output variables. The code uses NumPy arrays to store the data and computes a best-fit line, though the plotting itself is not shown in the provided snippet. The visualization would typically be rendered with a library like Matplotlib, displaying the data points and the fitted line to illustrate prediction accuracy. The dataset is explicitly defined in the file, with separate training and testing sets.

Aamirothman
48% match
Loading thumbnail…

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.

Aakeshavan
48% match