scprep v0.10.0

Changeset:

• removed unmaintained MATLAB code, moved python code to root directory
• all subsetting and filtering functions now accept extra_data, which allows the user to filter multiple array-likes at once (e.g., library sizes, sample labels, other metadata)
• split out row and column selection functionality from scprep.utils to scprep.select
• renamed some functions in scprep.filter for consistency
• significant additional functionality in scprep.plot.scatter
• added scprep.plot.marker_plot
• minor bugfixes

scprep v0.9.0

Changeset:

• removed Python 2.7 compatibility in line with matplotlib
• added plot module ported from PHATE
• implemented random projections sparse input PCA
• bugfix knnDREMI
• temporarily enforce pandas!=0.24.0 due to incompatibilities

scprep v0.8.0

Changeset:

• add scprep.filter.remove_duplicates
• add option to return library size from filtering and normalization functions
• add axis labels to histograms
• bugfix scprep.stats.knnDREMI and scprep.plot.show

scprep v0.7.0

Changeset:

• added scprep.stats for knn DREMI, mutual information and EMD
• HDF5 functionality now natively supports h5py and tables and is available in scprep.hdf5
• library size measurmenet and various filtering functions no longer coerce a pandas sparsedataframe to dense

scprep v0.6.0

Changeset:

• scprep.transform no longer densifies pandas.SparseDataFrame
• added scprep.reduce.pca for PCA applied to sparse data
• additional flexibility in scprep.plot passing keyword arguments to matplotlib.pyplot.hist
• added optional h5py backend for scprep.io.load_10X_HDF5 as an alternative to pytables
• minor bugfixes

scprep v0.5.0

Changes:

• gene_set_expression is now normalized by library size by default
• filter_library_size and filter_gene_set_expression can be done sample-wise if using percentile and providing sample labels.

scprep v0.4.0

Changes:

• added utils.combine_batches
• added plot.histogram
• fixed io.load_10X_HDF5
• io.load_fcs now returns three objects: FCS metadata (channel_meta), meta measurements (cell_meta), and genuine measurements (data).
• various bugfixes

scprep v0.3.1

scprep v0.3.1 includes mainly bugfixes.

Additional functionality:

• scprep.utils.get_gene_set and scprep.utils.get_cell_set gets lists of columns and indices from pandas DataFrames or lists of names using prefix, suffix, and regex filters.

Documented release

scprep now has documentation and a small tutorial.

scprep initial release

scprep offers a range of tools for loading, filtering, normalizing and transforming single-cell data, and is compatible with numpy, scipy and pandas matrices.