CHANGELOG.md

data_tools version history

v0.0.9 (WIP):

• In databases module:
  • Added option to retrieve gene symbols instead of UniProt AC in op_kinase_substrate function.
• In diffusion module:
  • Function coef_mat_hetero: Builds a block tri-diagonal coefficient matrix for a n-dimensional diffusion problem with heterogeneous diffusion coefficients.
• In models module:
  • Class Linear: Linear regression model using least squares.
  • Class PowerLaw: Fits a power law model to the provided data.
• In plots module:
  • Added option to show labels on significant points of volcano.
  • Function phase_portrait: Generates a phase portrait of a ODE system given the nullclines.
  • Function chordplot: Generates a chord plot from a collection of nodes and edges (and their sizes).
  • Function PCA: Computes the principal component analysis (PCA) and plots the results.

v0.0.8:

• Added signal module:
  • Function fconvolve: Convolves two vectors or arrays using Fast Fourier Transform (FFT).
  • Function gauss_kernel: Returns a N-dimensional Gaussian kernel.
  • Function gauss_noise: Applies additive Gaussian (white) noise to a given signal.
• In spatial module:
  • Function equidist_polar: For a given number of points (and optionally radius), returns the Cartesian coordinates of such number of equidistant points (in polar coordinates).
• In plots module:
  • Function cluster_hmap: Generates a heatmap with hierarchical clustering dendrograms attached.
  • Function upset_wrap: Wrapper for UpSetPlot package. Mostly just generates the Boolean multi-indexed pandas.Series the upsetplot.plot function needs as input.
  • Function similarity_heatmap: Given a group of sets, generates a heatmap with the similarity indices across each possible pair.
  • Function similarity_histogram: Given a group of sets, generates a histogram of the similarity indices across each possible pair (same-element pairs excluded).

v0.0.7:

• Support for Python 3.x versions.
• test_data_tools has tests for almost all classes and functions. Functions that include plots are skipped (cannot be tested). Tests missing are models.Lasso and spatial module (or maybe more).
  • Test suite has been improved and continuous integration with TravisCI has been implemented.
• Added spatial module:
  • Function neighbour_count: Given an array (up to three dimensions), returns another array with the same shape containing the counts of cells' neighbours whose value is zero.
  • Function get_boundaries: Given an array, returns either the mask where the boundary edges are or their counts if specified.
• In iterables module:
  • Function similarity: Computes the similarity index between two sets.
• In diffusion module:
  • Function build_mat unifies coefficient matrix building for any number of dimensions, numerical method and boundary condition.
  • ~~Function euler_explicit2D: Computes diffusion on a 2D space over a time-step using Euler explicit method.
  • Function euler_implicit_coef_mat: Computes the coefficient matrix to solve the diffusion problem with the Euler implicit method.
  • Function crank_nicolson_coef_mats: Computes the coefficient matrices to solve the diffusion problem with the Crank-Nicolson method.
  • Function build_coef_mat: Builds a coefficient matrix according to the central and neighbor coefficients, system size and boundary conditions.~~
• In models module:
  • Class DoseResponse improved and bugs fixed.
• In databases module:
  • Function op_kinase_substrate now can also download dephosphorylation interactions.
• In plots module:
  • Function venn now can plot global percentages instead of absolute counts. Also added the option to show the set sizes on the legend.
  • Function density now can handle data frames with samples on the columns.
  • Created 3 custom colormaps: cmap_bkgr (black-green), cmap_bkrd (black-red) and cmap_rdbkgr (red-black-green).

v0.0.6:

• Renamed sets module to iterables.
• In models module:
  • Class DoseResponse: Wrapper class for scipy.optimize.least_squares to fit dose-response curves on a pre-defined Hill function.
• In sets module:
  • Function unzip_dicts: Unzips the keys and values for any number of dictionaries passed as arguments (see below for examples).
  • Function chunk_this: For a given list L, returns another list of n-sized chunks from it (in the same order).
• In plots module:
  • Function venn now accepts up to 5 sets.
• In databases module:
  • Function op_kinase_substrate: Queries OmniPath to retrieve the kinase-substrate interactions for a given organism.
  • Function kegg_link: Queries a request to the KEGG database to find related entries using cross-references.
  • Function kegg_pathway_mapping: Makes a request to KEGG pathway mapping tool according to a given pathway ID. The user must provide a query of IDs to be mapped with their corresponding background colors (and optionally also foreground colors).
• Added diffusion module:
  • Function euler_explicit1D: Computes diffusion on a 1D space over a time-step using Euler explicit method.
• Renamed databases.up_query to databases.up_map
• Python 3 compatibility

v0.0.5:

• Added databases module:
  • Function up_query: Queries a request to UniProt.org in order to map a given list of identifiers.
• Linked GH-pages to HTML documentation.
• Added changelog.
• Implemented unit test.

v0.0.4

• In plots module:
  • Function density: Generates a density plot of the values on a data frame (row-wise).
  • Function venn: Plots a Venn diagram from a list of sets N. Number of sets must be between 2 and 4 (inclusive).
• In sets module:
  • Function subsets: Function that computes all possible logical relations between all sets on a list N and returns all subsets. This is, the subsets that would represent each intersecting area on a Venn diagram.
  • Function multi_union: DEPRECATED.

v0.0.3

• Added models module:
  • Class Lasso: Wrapper class inheriting from sklearn.linear_model.LogisticRegressionCV with L1 regularization.
    • Function fit_data: Fits the data to the logistic model.
    • Function plot_score: Plots the mean score across all folds obtained during CV. The optimum C parameter chosen and its score are highlighted.
    • Function plot_coef: Plots the non-zero coefficients for the fitted predictor features.
• In plots module:
  • Function piano_consensus: Generates a GSEA consensus score plot like R package piano's consensusScores function, but prettier. The main input is assumed to be a pandas.DataFrame whose data is the same as the rankMat from the result of consensusScores.

v0.0.2

• Documentation also available in PDF.
• Added plots module:
  • Function volcano: Generates a volcano plot from the differential expression data provided.

v0.0.1

• Automatic documentation implemented (HTML generated).
• Package can now be installed.
• Added sets module:
  • Function bit_or: Returns the bit operation OR between two bit-strings a and b. NOTE: a and b must have the same size.
  • Function find_min: Finds and returns the subset of vectors whose sum is minimum from a given set A.
  • Function in_all: Checks if a vector x is present in all sets contained in a list N.
  • Function multi_union: Returns the union set of all sets contained in a list N.
• Added strings module:
  • Function is_numeric: Determines if a string can be considered a numeric value. NaN is also considered, since it is float type.
  • Function join_str_lists: Joins element-wise two lists (or any 1D iterable) of strings with a given separator (if provided). Length of the input lists must be equal.