Skip to content

Instantly share code, notes, and snippets.

@daletovar
Last active February 24, 2021 20:18
Show Gist options
  • Save daletovar/874a2aa2579514ca73ec5f6ef71d8a46 to your computer and use it in GitHub Desktop.
Save daletovar/874a2aa2579514ca73ec5f6ef71d8a46 to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Leveraging sparsity for efficient data manipulation: the GCXS sparse array format"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Abstract:"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"identity (99% sparse) | symmetric (99% sparse) | random (99% sparse) \n",
"- | - | -\n",
"![alt](eye_matrix.png) | ![alt](symmetric2_matrix.png) | ![alt](random2.png) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Across many areas of science, sparsity is ubiquitous. Whether in the form of sparse data arrays, sparse linear operators, or the results from a sparsity-promoting optimization, sparse vectors, matrices, and tensors are common place in scientific computing. Sparse data structures, by definition, consist mostly of the same repeated value, typically zero. In the three matricies shown above, only 1% of the entries have a nonzero value. This feature can be exploited so as to reduce storage and runtime; sparse data is highly compressible because we can get away with only storing the nonzero values. There are many ways to store the nonzeros of a sparse data structure, which are manifested in many different sparse storage formats and are effecient at different kinds of operations. For example, a hashmap is well-suited for storing pairs of indices and their corresponding nonzero values and allows for efficient writing and mutating, but is poorly-suited for mathematical operations. By contrast, the compressed sparse row (CSR) storage format is well-suited for fast mathematical operations and compresses nicely but offers extremely inefficient writing and mutating capabilities. There are many widely-used sparse matrix storage formats, many of which, however, do not generalize well to higher dimensional arrays. Some are listed below:\n",
"\n",
" - dictionary of keys (DOK)\n",
" - compressed sparse column (CSC)\n",
" - compressed sparse row (CSR)\n",
" - coordinate list (COO)\n",
" - list of lists (LIL)\n",
" - block compressed sparse row (BSR)\n",
" \n",
"\n",
"In the scientific Python ecosystem, scipy.sparse has long been the de-facto sparse data module. It includes a full suite of sparse matrix formats and includes sparse linear system and eigen solvers. Additionaly, libraries that utilize sparse data such as scikit-learn rely on scipy.sparse. While being a mature and fast codebase, scipy.sparse emulates the numpy.matrix interface, which is restricted to two dimensions and is pending deprecation. As a result, scipy.sparse doesn't work well in places that numpy.ndarray does and it doesn't play nicely with the many libraries that emulate the numpy.ndarray interface such as dask and xarray. Since some of these libraries play important roles in scaling out the scientific Python ecosystem to larger data and distributed computing platforms, the reliance on the numpy.matrix interface holds scipy.sparse back.\n",
"\n",
"In efforts to address these concerns, the pydata/sparse library implements n-dimensional sparse data structures that emulate the numpy.ndarray interface. Pydata/sparse does this by extending two sparse matrix formats: the dictionary of keys (DOK) and the coordinate (COO) formats. DOK utilizes a hashmap to store the indices and values and is used mostly for writing. COO uses an array (ndim, nnz) to store the nonzero coordinates of each dimension and a seperate array to store the nonzero values. It implements element-wise operations, reductions, dot and tensordot operations, and is compatible with xarray and dask. The COO format, while fundamental, has several shortcomings in the way of compression and interoperability with the current users of scipy.sparse. Regarding compression, since COO stores a coordinate for each dimension of an array along with the nonzeros, as the number of dimensions increases, the storage requirements can quickly become unfavorable (O(ndim * nnz)). Also, most of the functions written inside scikit-learn and elsewhere work with compressed sparse row/column (CSR/CSC) matrices, not COO.\n",
"\n",
"With these issues in mind, pydata/sparse recently released the GCXS format, inspired by the GCRS/GCCS formats of [Shaikh and Hasan 2015](https://ieeexplore.ieee.org/document/7237032/authors#authors) . GCXS is an n-dimensional generalization of CSR/CSC. A main feature of it is that 2-d GCXS arrays are identical to CSR/CSC matrices (depending on which axis is compressed). This makes it so that libraries that use the CSR/CSC model internally, like scikit-learn, can easily be modified to work with GCXS arrays. GCXS is also capable of a high degree of compression."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can use neuroimaging data as an interesting test case. MRI data is collected in a 3-dimensional grid and is often transformed into a standard space. The standard Montreal Neurological Institute space (MNI152) has the dimensions (91, 109, 91). A brain in this space, however, typically occupies only a third of the full space, with the remainder of the array filled with zeros. When working with more dimensions like time or subjects, these datasets can get quite large and be computationally expensive. Because the data exceeds two dimensions, we can't use Scipy.sparse. When the density of an array grows too high, sparse array formats become less and less beneficial. Data with this high of a density (.33) pushes the limits of what is useful for most sparse array formats. Let's import what we need and see how well we can do."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import sparse\n",
"import numpy\n",
"import pandas\n",
"import nibabel\n",
"import scipy.sparse\n",
"import nilearn.plotting\n",
"import warnings\n",
"warnings.filterwarnings('ignore')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
" Here is a dataset with sixteen subjects:"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"img = nibabel.load('test_FA.nii.gz')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As a quick data check, let's have a look at the fractional anisotropy (FA) values of our first subjects' brain."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<nilearn.plotting.displays.OrthoSlicer at 0x7f41d725fcd0>"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 475.2x187.2 with 4 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"nilearn.plotting.plot_anat(img.slicer[...,0])"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table><tbody><tr><th style=\"text-align: left\">Format</th><td style=\"text-align: left\">gcxs</td></tr><tr><th style=\"text-align: left\">Data Type</th><td style=\"text-align: left\">float64</td></tr><tr><th style=\"text-align: left\">Shape</th><td style=\"text-align: left\">(91, 109, 91, 16)</td></tr><tr><th style=\"text-align: left\">nnz</th><td style=\"text-align: left\">4597231</td></tr><tr><th style=\"text-align: left\">Density</th><td style=\"text-align: left\">0.31832229797624495</td></tr><tr><th style=\"text-align: left\">Read-only</th><td style=\"text-align: left\">True</td></tr><tr><th style=\"text-align: left\">Size</th><td style=\"text-align: left\">70.1M</td></tr><tr><th style=\"text-align: left\">Storage ratio</th><td style=\"text-align: left\">0.6</td></tr><tr><th style=\"text-align: left\">Compressed Axes</th><td style=\"text-align: left\">(3,)</td></tr></tbody></table>"
],
"text/plain": [
"<GCXS: shape=(91, 109, 91, 16), dtype=float64, nnz=4597231, fill_value=0.0, compressed_axes=(3,)>"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data = sparse._compressed.GCXS.from_numpy(img.get_data(), compressed_axes=[3])\n",
"data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Looking at the storage ratio (.6), we can see the storage requirements of this GCXS array are 60% of the resources required to store this same data in a numpy.ndarray. So even with the high density (.318) we are able to get substantial compression with GCXS. Because of the high density and the high number of dimensions, COO is much less well-suited to store this data. "
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table><tbody><tr><th style=\"text-align: left\">Format</th><td style=\"text-align: left\">coo</td></tr><tr><th style=\"text-align: left\">Data Type</th><td style=\"text-align: left\">float64</td></tr><tr><th style=\"text-align: left\">Shape</th><td style=\"text-align: left\">(91, 109, 91, 16)</td></tr><tr><th style=\"text-align: left\">nnz</th><td style=\"text-align: left\">4597231</td></tr><tr><th style=\"text-align: left\">Density</th><td style=\"text-align: left\">0.31832229797624495</td></tr><tr><th style=\"text-align: left\">Read-only</th><td style=\"text-align: left\">True</td></tr><tr><th style=\"text-align: left\">Size</th><td style=\"text-align: left\">175.4M</td></tr><tr><th style=\"text-align: left\">Storage ratio</th><td style=\"text-align: left\">1.6</td></tr></tbody></table>"
],
"text/plain": [
"<COO: shape=(91, 109, 91, 16), dtype=float64, nnz=4597231, fill_value=0.0>"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.tocoo()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"With a storage ratio of 1.6, using COO requires us to use 160% of the storage requirements of a numpy.ndarray with this same data.\n",
"\n",
"One common need in a diffusion-weighted imaging analysis is to separate the white matter from the gray matter so that we can further examine the white matter. The goal is to define a white matter mask that we can use on all of our subjects. One simple way to do this is to take the mean accross subjects and threshold the data. When we are working with fractional anisotropy values (FA), the white matter and gray matter naturally seperate with higher values in the white matter. Since we are interested in computing across subjects it makes sense to compress the subjects dimension and group all of the spatial dimensions together. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Just like if this were a NumPy array, we can take the mean along the third axis, corresponding to subjects:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"mean = data.mean(axis=3)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can project the mean back onto a brain for visualization."
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<nilearn.plotting.displays.OrthoSlicer at 0x7f41cf1aafa0>"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 475.2x187.2 with 4 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"header = img.header\n",
"header['cal_min'] = mean.min()\n",
"header['cal_max'] = mean.max()\n",
"mean_dense = mean.todense()\n",
"mean_img = nibabel.Nifti1Image(mean_dense, header=header, affine=img.affine)\n",
"nilearn.plotting.plot_stat_map(mean_img, colorbar=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The white matter comes out very bright. We can threshold the data to fully separate the white matter from the gray."
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<nilearn.plotting.displays.OrthoSlicer at 0x7f41ceec3160>"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 475.2x187.2 with 4 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"mean_dense[mean_dense < .2] = 0\n",
"mean_img = nibabel.Nifti1Image(mean_dense, header=header, affine=img.affine)\n",
"nilearn.plotting.plot_stat_map(mean_img, colorbar=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In this short demonstration we were able to use Sparse arrays instead of Numpy arrays without needing to learn a different interface, since Sparse copies the Numpy interface. Instead of computing the mean across every measurement we were able to compute over only the necessary measurements, while using only 60% of the memory required by Numpy."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Pydata/sparse is primarily focused on providing sparse array objects. The objects are compatible with the algorithms in Scipy.sparse, which include direct and iterative sparse linear system solvers and iterative eigen solvers. \n",
"\n",
"The singular value decomposition is a critical procedure for a lot of data analytics applications. Using sparse arrays alongside scipy makes this procedure feel natural. Similar to generating \"eigen faces,\" with the SVD we can look at the low-rank structure of our data and visualize \"eigen brains,\" possibly to denoise the data or examine structure, but mostly for fun.\n",
"\n",
"The singular value decomposition requires that we fold the sparse into a new shape. Otherwise we raise an error."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"try: scipy.sparse.linalg.svds(data, 15)\n",
"except ValueError: ..."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"A singular value decomposition can now operate on the folded 2-D form."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"folded = sparse._compressed.GCXS(\n",
" (data.data, data.indices, data.indptr),\n",
" shape=data._compressed_shape[::-1], \n",
" compressed_axes=[1])\n",
"\n",
"U, S, Vᵀ = scipy.sparse.linalg.svds(folded, 15)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The columns of U hold the dominant spatial patterns of the data. We can reshape these vectors and project them onto brains:"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 525.6x187.2 with 5 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 525.6x187.2 with 5 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 525.6x187.2 with 5 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"for i in range(3):\n",
" image = U[:,i].reshape((91,109,91))\n",
" header['cal_min'], header['cal_max'] = image.min(), image.max()\n",
" brain = nibabel.Nifti1Image(image, header=header, affine=img.affine)\n",
" nilearn.plotting.plot_stat_map(brain)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Conclusion"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Problems involving sparsity are pervasive in scientific computing. Efficiently storing and manipulating large sparse arrays requires thoughtful storage formats. In an ecosystem where more and more libraries emulate and interact with NumPy, it's important that sparse arrays be compatable, especially with libraries like Dask, for predictablity of behavior and scalability. Pydata/sparse arrays offer these features. Each storage format offers different advantages. Because GCXS is identical to CSR/CSC for two-dimensional arrays, it is primed to work effectively in all the libraries that expect CSR/CSC. It also excels in the same areas as CSR/CSC; namely in fast matrix multiplication. Above and beyond this, GCXS offers good compression for high-dimensional arrays."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Technical Details"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Understanding how GCXS works requires first understanding how the CSR/CSC model works for matrices. For a CSR matrix, we store three arrays: the nonzero data, the indices corresponding the columns of each of the each the nonzero data, and an array of integers that point to where each row begins in the indices array.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Here we make a 5x5 CSR matrix"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"a = sparse.random((5,5), density=.2, format='gcxs')"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table><tbody><tr><th style=\"text-align: left\">Format</th><td style=\"text-align: left\">gcxs</td></tr><tr><th style=\"text-align: left\">Data Type</th><td style=\"text-align: left\">float64</td></tr><tr><th style=\"text-align: left\">Shape</th><td style=\"text-align: left\">(5, 5)</td></tr><tr><th style=\"text-align: left\">nnz</th><td style=\"text-align: left\">5</td></tr><tr><th style=\"text-align: left\">Density</th><td style=\"text-align: left\">0.2</td></tr><tr><th style=\"text-align: left\">Read-only</th><td style=\"text-align: left\">True</td></tr><tr><th style=\"text-align: left\">Size</th><td style=\"text-align: left\">128</td></tr><tr><th style=\"text-align: left\">Storage ratio</th><td style=\"text-align: left\">0.6</td></tr><tr><th style=\"text-align: left\">Compressed Axes</th><td style=\"text-align: left\">(0,)</td></tr></tbody></table>"
],
"text/plain": [
"<GCXS: shape=(5, 5), dtype=float64, nnz=5, fill_value=0.0, compressed_axes=(0,)>"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[0.25710831, 0. , 0.56822 , 0. , 0. ],\n",
" [0. , 0. , 0. , 0.54305051, 0. ],\n",
" [0. , 0.37660311, 0. , 0. , 0.75337383],\n",
" [0. , 0. , 0. , 0. , 0. ],\n",
" [0. , 0. , 0. , 0. , 0. ]])"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a.todense()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As we can see, the data is sorted in row-major ordering:"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0.25710831, 0.56822 , 0.54305051, 0.37660311, 0.75337383])"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a.data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The indices store which columns have nonzeros:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0, 2, 3, 1, 4])"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a.indices"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The index pointer array tells us which rows are paired with the columns in 'indices'"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0, 2, 3, 5, 5, 5])"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a.indptr"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The columns stored in 'indices' for the i-th row correspond to a.indices[a.indptr[i]:a.indptr[i+1]]"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The nonzero columns of row 0 [0 2]\n",
"The nonzero columns of row 1 [3]\n",
"The nonzero columns of row 2 [1 4]\n",
"The nonzero columns of row 3 []\n",
"The nonzero columns of row 4 []\n"
]
}
],
"source": [
"for i in range(5):\n",
" print(\"The nonzero columns of row %s \" %(i), a.indices[a.indptr[i]:a.indptr[i+1]])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The CSC data structure uses the same idea but stores the data in column-major ordering:"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table><tbody><tr><th style=\"text-align: left\">Format</th><td style=\"text-align: left\">gcxs</td></tr><tr><th style=\"text-align: left\">Data Type</th><td style=\"text-align: left\">float64</td></tr><tr><th style=\"text-align: left\">Shape</th><td style=\"text-align: left\">(5, 5)</td></tr><tr><th style=\"text-align: left\">nnz</th><td style=\"text-align: left\">5</td></tr><tr><th style=\"text-align: left\">Density</th><td style=\"text-align: left\">0.2</td></tr><tr><th style=\"text-align: left\">Read-only</th><td style=\"text-align: left\">True</td></tr><tr><th style=\"text-align: left\">Size</th><td style=\"text-align: left\">128</td></tr><tr><th style=\"text-align: left\">Storage ratio</th><td style=\"text-align: left\">0.6</td></tr><tr><th style=\"text-align: left\">Compressed Axes</th><td style=\"text-align: left\">(1,)</td></tr></tbody></table>"
],
"text/plain": [
"<GCXS: shape=(5, 5), dtype=float64, nnz=5, fill_value=0.0, compressed_axes=(1,)>"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b = a.change_compressed_axes([1])\n",
"b"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[0.25710831, 0. , 0.56822 , 0. , 0. ],\n",
" [0. , 0. , 0. , 0.54305051, 0. ],\n",
" [0. , 0.37660311, 0. , 0. , 0.75337383],\n",
" [0. , 0. , 0. , 0. , 0. ],\n",
" [0. , 0. , 0. , 0. , 0. ]])"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b.todense()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now the data is sorted in column-major form:"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0.25710831, 0.37660311, 0.56822 , 0.54305051, 0.75337383])"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b.data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The 'indices' array now stores the rows of the nonzero values:"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0, 2, 0, 1, 2])"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b.indices"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"And now the index pointer array tells us which of the rows in 'indices' correspond to what column:"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0, 1, 2, 3, 4, 5])"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b.indptr"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The nonzero rows of column 0 [0]\n",
"The nonzero rows of column 1 [2]\n",
"The nonzero rows of column 2 [0]\n",
"The nonzero rows of column 3 [1]\n",
"The nonzero rows of column 4 [2]\n"
]
}
],
"source": [
"for i in range(5):\n",
" print(\"The nonzero rows of column %s \" %(i), b.indices[b.indptr[i]:b.indptr[i+1]])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For GCXS, CSR and CSC can be thought of as the same data structure, but with a different compressed axis. For CSR the first axis (rows) are compressed whereas for CSC the second axis (columns) are compressed."
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(0,)"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a.compressed_axes"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(1,)"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b.compressed_axes"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The way we can work with arrays of higher dimensions using this kind of scheme is by transposing axes and reshaping the original array into a two-dimensional array and applying the same methods. We transpose the axes to put the compressed axes first and the reshape the array to two dimensions with the resulting shape: (product of compressed axes, product of uncompressed axes). For a 3x4x5 array, if we compress the first axis then we will store a CSR matrix with a shape of (3, 4x5=20)."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"x = sparse.random((3,4,5), density=.2, format='gcxs')"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(3, 20)"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"x._compressed_shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If we compress the second axis, then we transpose the array and reshape it into a (4, 3x5=15) matrix."
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(4, 15)"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"x.change_compressed_axes([1])._compressed_shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can also compress multiple axes:"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(15, 4)"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"x.change_compressed_axes([0,2])._compressed_shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![alt](FA_matrix.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Different choices in compressed axes affects the degree of compression and the speed of certain operations, especially indexing. Like CSR/CSC, the storage demands are determined by the size of the compressed axes and the number of nonzeros. Because the storage demands of a GCXS array are not dependent on the number of dimensions of the array, GCXS compresses well even with high dimensional arrays. The best compression is achieved by compressing the smallest axis:"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"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>compressed_axes</th>\n",
" <th>percent_compression</th>\n",
" <th>compressed_shape</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>[0]</td>\n",
" <td>30.003429</td>\n",
" <td>(5, 35000)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>[1]</td>\n",
" <td>30.012000</td>\n",
" <td>(20, 8750)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>[2]</td>\n",
" <td>30.020571</td>\n",
" <td>(35, 5000)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>[3]</td>\n",
" <td>30.029143</td>\n",
" <td>(50, 3500)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>[0, 1]</td>\n",
" <td>30.057714</td>\n",
" <td>(100, 1750)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>[0, 2]</td>\n",
" <td>30.100571</td>\n",
" <td>(175, 1000)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>[0, 3]</td>\n",
" <td>30.143429</td>\n",
" <td>(250, 700)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>[1, 2]</td>\n",
" <td>30.400571</td>\n",
" <td>(700, 250)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>[1, 3]</td>\n",
" <td>30.572000</td>\n",
" <td>(1000, 175)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>[2, 3]</td>\n",
" <td>31.000571</td>\n",
" <td>(1750, 100)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>[0, 1, 2]</td>\n",
" <td>32.000571</td>\n",
" <td>(3500, 50)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11</th>\n",
" <td>[0, 2, 3]</td>\n",
" <td>35.000571</td>\n",
" <td>(8750, 20)</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12</th>\n",
" <td>[1, 2, 3]</td>\n",
" <td>50.000571</td>\n",
" <td>(35000, 5)</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" compressed_axes percent_compression compressed_shape\n",
"0 [0] 30.003429 (5, 35000)\n",
"1 [1] 30.012000 (20, 8750)\n",
"2 [2] 30.020571 (35, 5000)\n",
"3 [3] 30.029143 (50, 3500)\n",
"4 [0, 1] 30.057714 (100, 1750)\n",
"5 [0, 2] 30.100571 (175, 1000)\n",
"6 [0, 3] 30.143429 (250, 700)\n",
"7 [1, 2] 30.400571 (700, 250)\n",
"8 [1, 3] 30.572000 (1000, 175)\n",
"9 [2, 3] 31.000571 (1750, 100)\n",
"10 [0, 1, 2] 32.000571 (3500, 50)\n",
"11 [0, 2, 3] 35.000571 (8750, 20)\n",
"12 [1, 2, 3] 50.000571 (35000, 5)"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"y = sparse.random((5,20,35,50), density=.15, format='gcxs')\n",
"axes = ([0], [1], [2], [3], [0,1], \n",
" [0,2], [0,3], [1,2], \n",
" [1,3], [2,3], [0,1,2],\n",
" [0,2,3], [1,2,3])\n",
"percent_stored = []\n",
"shapes = []\n",
"for axis in axes:\n",
" y = y.change_compressed_axes(axis)\n",
" percent_stored.append(y.nbytes/(np.prod(y.shape)*y.dtype.itemsize) *100)\n",
" shapes.append(y._compressed_shape)\n",
" #print('compressed axes: %s compression ratio: %s ' %(axis, y.nbytes/(np.prod(y.shape)*y.dtype.itemsize) *100),\n",
" # 'shape of underlying matrix: ', y._compressed_shape)\n",
"df = pandas.DataFrame()\n",
"df['compressed_axes'] = axes\n",
"df['percent_compression'] = percent_stored\n",
"df['compressed_shape'] = shapes\n",
"df"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As we can see, compressing the three largest axes together gets the worst compression. Nonetheless, in this example the worst compression still results in storing only half of the bytes needed for the equivelant dense array. Also, with this number of dimensions, the compression will usually be much better than COO:"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"75.0"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"y.tocoo().nbytes/(np.prod(y.shape)*y.dtype.itemsize) *100"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"GCXS arrays compress well and look and behave like NumPy arrays. GCXS currently supports the following operations:\n",
" - indexing\n",
" - multiplication\n",
" - dot\n",
" - tensordot\n",
" - matmul\n",
" - element-wise operations (including broadcasting)\n",
" - np.sin\n",
" - np.sqrt\n",
" - np.add\n",
" - np.log\n",
" - ...\n",
" - reductions\n",
" - min\n",
" - max\n",
" - sum\n",
" - prod\n",
" - mean\n",
" - std\n",
" - var\n",
" - ..."
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Display the source blob
Display the rendered blob
Raw
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment