Sctriangulate

In this new release:

1. a ready-to-use python package
2. can generate publication-quality figure
3. passed all the test functions
4. make sure all the default argument values are correct

In this new version:

1. better inspection visualization
2. add 7 modes to exclude artifact genes
3. filter genes for the heatmap visualization

First ready version, you can run the tool through the command line.

Four modes:

metrics
shapley
combine
cluster

In this new version:

1. completely parallelizable, no matter how many cells you have and how many set of annotations, it is readily scalable (no need to tune the cores, the program figure out by itself)
2. accurately garbage collection and memory management, try best to avoid the subprocess consume too much memory
3. throwing intermediate file to re-run and debug
4. in case memory overhead in a parallel version, I add scTriangulate sequential version

In this new version:

1. tested it is able to apply to human datasets
2. make sure your adata.raw is a sparse matrix (which usually is)
3. fix the bug that in the reassign score step, if n_sample/centroid is less than n_neighbors, set neighbors = n_centroid
4. fix the bug that when filter any cluster whose cell number = 1 at the first step, change 'gs' to key

In this small update version:

1. precomputing the cluster size, reducing repetitive computing in tie-breaking and pruning
2. additional pruning mechanisms, proportion_to_self must be greater than 0.1 as well even if neraly=True
3. add listening log file while computing Shapley, flush to a log file