Skip to content

plots.md

Latest commit

 

History

History
279 lines (197 loc) · 8.87 KB

plots.md

File metadata and controls

279 lines (197 loc) · 8.87 KB

Visualization: Plots

The metainformant.visualization.plots package provides the core plotting functions organized into five submodules: basic charts, general biological plots, specialized diagrams, multi-dimensional visualizations, and animations.

All plot functions return a matplotlib.axes.Axes object and accept an optional output_path parameter to save the figure to disk.

Submodule Overview

Submodule Description Plot Types
basic Fundamental chart types built on matplotlib line, scatter, heatmap, bar, pie, area, step
general Biology-oriented statistical plots volcano, manhattan, QQ, PCA, expression heatmap, correlation
specialized Advanced diagram types for systems biology Venn, Sankey, chord, alluvial, circular bar, UpSet
multidim Multi-dimensional data exploration pairwise, parallel coordinates, radar, 3D scatter
animations Animated visualizations for temporal/dynamic data time series, evolution, clustering, network, trajectory

Basic Plots (plots.basic)

lineplot

import numpy as np
from metainformant.visualization import lineplot

x = np.linspace(0, 10, 100)
y = np.sin(x)
ax = lineplot(x, y, label="sine wave", color="blue")
Parameter Type Default Description
x np.ndarray required X-axis data
y `np.ndarray None` None
ax `Axes None` None
output_path `str Path None`

scatter_plot

from metainformant.visualization import scatter_plot

ax = scatter_plot(np.random.randn(100), np.random.randn(100), alpha=0.6)

heatmap

from metainformant.visualization import heatmap

data = np.random.rand(10, 10)
ax = heatmap(data, cmap="viridis")

Requires a 2D numpy array. Includes an automatic colorbar.

bar_plot

from metainformant.visualization import bar_plot

categories = np.array(["A", "B", "C", "D"])
values = np.array([10, 25, 15, 30])
ax = bar_plot(categories, values, color="teal")

pie_chart

from metainformant.visualization import pie_chart

sizes = np.array([35, 25, 20, 20])
ax = pie_chart(sizes, labels=["A", "B", "C", "D"], autopct="%1.1f%%")

area_plot

Filled area chart using matplotlib.fill_between.

from metainformant.visualization import area_plot

ax = area_plot(np.arange(50), np.random.rand(50).cumsum(), alpha=0.4)

step_plot

Step function plot.

from metainformant.visualization import step_plot

ax = step_plot(np.arange(20), np.random.randint(0, 10, 20))

General Plots (plots.general)

volcano_plot

Visualize differential expression results as log2 fold change vs. -log10 p-value.

import pandas as pd
from metainformant.visualization import volcano_plot

de_results = pd.DataFrame({
    "log2FoldChange": np.random.randn(1000),
    "padj": np.random.uniform(0, 1, 1000),
})
ax = volcano_plot(de_results, log2fc_col="log2FoldChange", pval_col="padj")

manhattan_plot

Genome-wide association study results plotted by chromosomal position.

from metainformant.visualization import manhattan_plot

gwas_data = pd.DataFrame({
    "CHR": np.repeat(range(1, 23), 50),
    "BP": np.random.randint(1, 1_000_000, 1100),
    "P": np.random.uniform(0, 1, 1100),
})
ax = manhattan_plot(gwas_data, pos_col="BP", pval_col="P")

qq_plot

Quantile-quantile plot for assessing distributional assumptions and identifying genomic inflation.

from metainformant.visualization import qq_plot

pvalues = np.random.uniform(0, 1, 500)
ax = qq_plot(pvalues)

expression_heatmap

Clustered heatmap of gene expression data. Uses seaborn if available, falls back to matplotlib.

from metainformant.visualization import expression_heatmap

expr_data = pd.DataFrame(
    np.random.randn(20, 6),
    index=[f"Gene_{i}" for i in range(20)],
    columns=[f"Sample_{i}" for i in range(6)],
)
ax = expression_heatmap(expr_data, cmap="RdBu_r")

pca_plot

Scatter plot of pre-computed principal components with variance annotations.

from metainformant.visualization import pca_plot

pca_data = pd.DataFrame({
    "PC1": np.random.randn(50),
    "PC2": np.random.randn(50),
    "PC1_variance": [25.5] * 50,
    "PC2_variance": [18.3] * 50,
    "group": np.random.choice(["Control", "Treatment"], 50),
})
ax = pca_plot(pca_data, hue="group")

correlation_heatmap

Correlation matrix as annotated heatmap.

from metainformant.visualization import correlation_heatmap

df = pd.DataFrame(np.random.randn(100, 5), columns=["A", "B", "C", "D", "E"])
ax = correlation_heatmap(df, annot=True)

Specialized Plots (plots.specialized)

Function Description
plot_venn_diagram 2-3 set Venn diagram with intersection counts
plot_sankey_diagram Flow diagram showing transitions between states
plot_chord_diagram Circular diagram for pairwise relationships
plot_alluvial_diagram Parallel category flow visualization
plot_circular_barplot Radial bar chart for cyclic data
plot_network_circular_layout Network graph with circular node arrangement
plot_upset_plot UpSet plot for multi-set intersection analysis 
from metainformant.visualization import plot_venn_diagram

sets = {
    "Gene Set A": {"BRCA1", "TP53", "EGFR", "MYC"},
    "Gene Set B": {"TP53", "KRAS", "MYC", "PIK3CA"},
}
ax = plot_venn_diagram(sets)

Multi-Dimensional Plots (plots.multidim)

Function Description
plot_pairwise_relationships Scatter matrix (pairplot) for all numeric columns
plot_parallel_coordinates Parallel coordinate plot for multivariate data
plot_radar_chart Radar/spider chart for comparing profiles
plot_3d_scatter 3D scatter plot using mpl_toolkits 
from metainformant.visualization import plot_pairwise_relationships

df = pd.DataFrame(np.random.randn(100, 4), columns=["W", "X", "Y", "Z"])
ax = plot_pairwise_relationships(df)

Animations (plots.animations)

All animation functions return a (Figure, FuncAnimation) tuple and can save to GIF via the output_path parameter.

Function Description
animate_time_series Progressive line plot revealing data over time
animate_evolution Sequence/population change across generations
animate_clustering Step-by-step clustering algorithm visualization
animate_network Dynamic network graph with edge/node changes
animate_trajectory Point movement path animation 
from metainformant.visualization import animate_time_series

data = np.random.randn(100, 3).cumsum(axis=0)
fig, anim = animate_time_series(data, interval=100, output_path="output/ts.gif")

Common Parameters

All plot functions share these optional parameters:

Parameter Type Description
ax `Axes None`
output_path `str Path
figsize tuple Figure size as (width, height) in inches
**kwargs Any Passed through to underlying matplotlib call

Figures are saved at 300 DPI with tight bounding boxes. Parent directories are created automatically.

See Also

  • metainformant.visualization.config.themes -- Theme and styling management
  • metainformant.visualization.config.palettes -- Color palettes (Wong, IBM, etc.)
  • metainformant.visualization.dashboards -- Composite multi-panel layouts
  • metainformant.visualization.interactive_dashboards -- Plotly-based interactive plots