PHATE v2.0.0

What's Changed

• Update LICENSE to Yale Ventures Non-Commercial License by @dbhaskar92 in #144
• removed s_gd2 dependency. now just using smacof by @MattScicluna in #156
• 158 update requirements and installation procedure by @bjoaofelipe in #160
• Add random landmarking by @MattScicluna in #161
• Add back sgd mds by @MattScicluna in #163
• Updated tutorial to remove scprep by @bjoaofelipe in #165
• changed lr schedule for SGD MDS by @MattScicluna in #164
• Add disconnection warning by @MattScicluna in #166
• Add pytests by @MattScicluna in #168
• updated version info and added contributors info by @MattScicluna in #167
• Added unittest for data types (scanpy, pandas, scipy) by @bjoaofelipe in #170

New Contributors

• @dbhaskar92 made their first contribution in #144
• @MattScicluna made their first contribution in #156
• @bjoaofelipe made their first contribution in #160

Full Changelog: v1.0.11...2.0.0

PHATE v0.3.0

Changeset:

• add cluster submodule
  • adds cluster.kmeans function to perform KMeans on the diffusion potential
• add dpi argument to plot.rotate_scatter3d

PHATE v0.2.10

Changeset:

• phate.plot.scatter now accepts a list of matplotlib colors
• phate.plot.scatter now returns the axis on which the plot was drawn
• minor logging changes

PHATE v0.2.9

Changes:

• Use tasklogger for logging; fixes issues with Windows 7 verbose output
• Use scprep in tutorials for simpler analysis
• Change default value of alpha to 40 for faster computation of kernel
• phate.plot.scatter now takes keyword arguments x, y and z rather than a list-like
• phate.plot.scatter has keyword arguments for axis label prefix label_prefix (as an alternative to labelling axes individually), plot title title, and legend location legend_loc
• minor bugfixes

PHATE v0.2.8

PHATE v0.2.8 adds plotting utilities:

• phate.plot.scatter2d: 2D scatterplot
• phate.plot.scatter3d: 3D scatterplot
• phate.plot.rotate_scatter3d: rotating 3D scatterplot (gif or mp4)

All plotting functions accept either data or a PHATE object as input. The color vector c can be continuous or categorical, and need not be numeric. Legends / colorbars are generated with randomized point order on the canvas.

Sparse fast alpha and gamma informational distances

PHATE now implements a sparse, fast alpha decay kernel which has minimal memory requirements, compared to the old alpha decay which required a parwise distance matrix. Alpha decay now runs by default. In order to use the k nearest neighbors kernel, run with a=None.

Other notable changes:

• graphtools implementation of kernel matrices
• gamma replaces potential_method for the selection of informational distances. gamma=1 is equivalent to a log potential (default) and gamma=0 is equivalent to a square root potential
• arguments alpha_decay and potential_method are deprecated.
• PHATE is now implemented as a visualization method in Scanpy: https://scanpy.readthedocs.io/

Accept affinity and distance matrices

The Python version of PHATE now accepts both distance matrices and affinity matrices with the keyword knn_dist='precomputed'.

We assume distance matrices have only zeroes along the diagonal, and affinity matrices have no zeroes on the diagonal.

Introduce scanpy support

PHATE now accepts scanpy's native AnnData format

PHATE 2.0 (Fast scalable PHATE)

Version 2.0 implements fast scalable PHATE in Python (2.7, >=3.5), MATLAB and R.

PHATE now runs in seconds to minutes on tens of thousands of cells. Benchmarking shows runtime of ~3 hours on >1,000,000 cells.

Key changes:

• t is automatically chosen using Von Neumann Entropy using t='auto' by default.
• n_landmark determines the number of landmarks to use for scalable PHATE. n_landmark=None ([] on MATLAB, NA in R) specifies exact PHATE.
• default parameters have changed: k=15, a=10 but is ignored if n_cells>=n_landmark.