Skip to content

Latest commit

 

History

History
610 lines (496 loc) · 20.7 KB

README.md

File metadata and controls

610 lines (496 loc) · 20.7 KB

Hero Plotille

Plotille

CI codecov Tested CPython Versions Tested PyPy Versions PyPi version Downloads PyPi license

Plots, scatter plots, histograms and heatmaps in the terminal using braille dots, and foreground and background colors - with no dependancies. Make complex figures using the Figure class or make fast and simple plots using graphing function - similar to a very small sibling to matplotlib. Or use the canvas to plot dots, lines and images yourself.

Install:

pip install plotille

Similar to other libraries:

  • like drawille, but focused on graphing – plus X/Y-axis.
  • like termplot, but with braille (finer dots), left to right histogram and linear interpolation for plotting function.
  • like termgraph (not on pypi), but very different style.
  • like terminalplot, but with braille, X/Y-axis, histogram, linear interpolation.

Basic support for timeseries plotting is provided with release 3.2: for any X or Y values you can also add datetime.datetime, pendulum.datetime or numpy.datetime64 values. Labels are generated respecting the difference of x_limits and y_limits.

Support for heatmaps using background colors for figures and displaying images binary with braille, or in color with background colors using the canvas - provided with release 4.0

If you are still using python 2.7, please use plotille v4 or before. With v5 I am dropping support for python 2.7, as the effort to maintain the discontinued version is too much.

Documentation

In [1]: import plotille
In [2]: import numpy as np
In [3]: X = np.sort(np.random.normal(size=1000))

Figure

To construct plots the recomended way is to use a Figure:

In [12]: plotille.Figure?
Init signature: plotille.Figure()
Docstring:
Figure class to compose multiple plots.

Within a Figure you can easily compose many plots, assign labels to plots
and define the properties of the underlying Canvas. Possible properties that
can be defined are:

    width, height: int    Define the number of characters in X / Y direction
                          which are used for plotting.
    x_limits: float       Define the X limits of the reference coordinate system,
                          that will be plottered.
    y_limits: float       Define the Y limits of the reference coordinate system,
                          that will be plottered.
    color_mode: str       Define the used color mode. See `plotille.color()`.
    with_colors: bool     Define, whether to use colors at all.
    background: multiple  Define the background color.
    x_label, y_label: str Define the X / Y axis label.

Basically, you create a Figure, define the properties and add your plots. Using the show() function, the Figure generates the plot using a new canvas:

In [13] fig = plotille.Figure()
In [14] fig.width = 60
In [15] fig.height = 30
In [16] fig.set_x_limits(min_=-3, max_=3)
In [17] fig.set_y_limits(min_=-1, max_=1)
In [18] fig.color_mode = 'byte'
In [19] fig.plot([-0.5, 1], [-1, 1], lc=25, label='First line')
In [20] fig.scatter(X, np.sin(X), lc=100, label='sin')
In [21] fig.plot(X, (X+2)**2 , lc=200, label='square')
In [22] print(fig.show(legend=True))

Example figure

The available plotting functions are:

# create a plot with linear interpolation between points
Figure.plot(self, X, Y, lc=None, interp='linear', label=None, marker=None)
# create a scatter plot with no interpolation between points
Figure.scatter(self, X, Y, lc=None, label=None, marker=None)
# create a histogram over X
Figure.histogram(self, X, bins=160, lc=None)
# print texts at coordinates X, Y
Figure.text(self, X, Y, texts, lc=None)

# The following functions use relative coordinates on the canvas
# i.e. all coordinates are \in [0, 1]
# plot a vertical line at x
Figure.axvline(self, x, ymin=0, ymax=1, lc=None)
# plot a vertical rectangle from (xmin,ymin) to (xmax, ymax).
Figure.axvspan(self, xmin, xmax, ymin=0, ymax=1, lc=None)
# plot a horizontal line at y
Figure.axhline(self, y, xmin=0, xmax=1, lc=None)
# plot a horizontal rectangle from (xmin,ymin) to (xmax, ymax).
Figure.axhspan(self, ymin, ymax, xmin=0, xmax=1, lc=None)

# Display data as an image, i.e. on a 2D regular raster.
Figure.imgshow(self, X, cmap=None)

Other interesting functions are:

# remove all plots, texts, spans and images from the figure
Figure.clear(self)
# Create a canvas, plot the registered plots and return the string for displaying the plot
Figure.show(self, legend=False)

Please have a look at the examples/ folder.

Graphing

There are some utility functions for fast graphing of single plots.

Plot

In [4]: plotille.plot?
Signature:
plotille.plot(
    X,
    Y,
    width=80,
    height=40,
    X_label='X',
    Y_label='Y',
    linesep=os.linesep,
    interp='linear',
    x_min=None,
    x_max=None,
    y_min=None,
    y_max=None,
    lc=None,
    bg=None,
    color_mode='names',
    origin=True,
    marker=None,
)
Docstring:
Create plot with X , Y values and linear interpolation between points

Parameters:
    X: List[float]         X values.
    Y: List[float]         Y values. X and Y must have the same number of entries.
    width: int             The number of characters for the width (columns) of the canvas.
    hight: int             The number of characters for the hight (rows) of the canvas.
    X_label: str           Label for X-axis.
    Y_label: str           Label for Y-axis. max 8 characters.
    linesep: str           The requested line seperator. default: os.linesep
    interp: Optional[str]  Specify interpolation; values None, 'linear'
    x_min, x_max: float    Limits for the displayed X values.
    y_min, y_max: float    Limits for the displayed Y values.
    lc: multiple           Give the line color.
    bg: multiple           Give the background color.
    color_mode: str        Specify color input mode; 'names' (default), 'byte' or 'rgb'
                           see plotille.color.__docs__
    origin: bool           Whether to print the origin. default: True
    marker: str            Instead of braille dots set a marker char for actual values.

Returns:
    str: plot over `X`, `Y`.

In [5]: print(plotille.plot(X, np.sin(X), height=30, width=60))

Example plot

Scatter

In [6]: plotille.scatter?
Signature:
plotille.scatter(
    X,
    Y,
    width=80,
    height=40,
    X_label='X',
    Y_label='Y',
    linesep='\n',
    x_min=None,
    x_max=None,
    y_min=None,
    y_max=None,
    lc=None,
    bg=None,
    color_mode='names',
    origin=True,
    marker=None,
)
Docstring:
Create scatter plot with X , Y values

Basically plotting without interpolation:
    `plot(X, Y, ... , interp=None)`

Parameters:
    X: List[float]       X values.
    Y: List[float]       Y values. X and Y must have the same number of entries.
    width: int           The number of characters for the width (columns) of the canvas.
    hight: int           The number of characters for the hight (rows) of the canvas.
    X_label: str         Label for X-axis.
    Y_label: str         Label for Y-axis. max 8 characters.
    linesep: str         The requested line seperator. default: os.linesep
    x_min, x_max: float  Limits for the displayed X values.
    y_min, y_max: float  Limits for the displayed Y values.
    lc: multiple         Give the line color.
    bg: multiple         Give the background color.
    color_mode: str      Specify color input mode; 'names' (default), 'byte' or 'rgb'
                         see plotille.color.__docs__
    origin: bool         Whether to print the origin. default: True
    marker: str          Instead of braille dots set a marker char.

Returns:
    str: scatter plot over `X`, `Y`.

In [7]: print(plotille.scatter(X, np.sin(X), height=30, width=60))

Example scatter

Hist

Inspired by crappyhist (link is gone, but I made a gist).

In [8]: plotille.hist?
Signature:
plotille.hist(
    X,
    bins=40,
    width=80,
    log_scale=False,
    linesep='\n',
    lc=None,
    bg=None,
    color_mode='names',
)
Docstring:
Create histogram over `X` from left to right

The values on the left are the center of the bucket, i.e. `(bin[i] + bin[i+1]) / 2`.
The values on the right are the total counts of this bucket.

Parameters:
    X: List[float]       The items to count over.
    bins: int            The number of bins to put X entries in (rows).
    width: int           The number of characters for the width (columns).
    log_scale: bool      Scale the histogram with `log` function.
    linesep: str         The requested line seperator. default: os.linesep
    lc: multiple         Give the line color.
    bg: multiple         Give the background color.
    color_mode: str      Specify color input mode; 'names' (default), 'byte' or 'rgb'
                         see plotille.color.__docs__

Returns:
    str: histogram over `X` from left to right.

In [9]: print(plotille.hist(np.random.normal(size=10000)))

Example hist

Hist (aggregated)

This function allows you to create a histogram when your data is already aggregated (aka you don't have access to raw values, but you have access to bins and counts for each bin).

This comes handy when working with APIs such as OpenTelemetry Metrics API where views such as ExplicitBucketHistogramAggregation only expose access to aggregated values (counts for each bin / bucket).

In [8]: plotille.hist_aggregated?
Signature:
plotille.hist_aggregated(
    counts,
    bins,
    width=80,
    log_scale=False,
    linesep='\n',
    lc=None,
    bg=None,
    color_mode='names',
)
Docstring:
Create histogram for aggregated data.

Parameters:
    counts: List[int]    Counts for each bucket.
    bins: List[float]    Limits for the bins for the provided counts: limits for
                         bin `i` are `[bins[i], bins[i+1])`.
                         Hence, `len(bins) == len(counts) + 1`.
    width: int           The number of characters for the width (columns).
    log_scale: bool      Scale the histogram with `log` function.
    linesep: str         The requested line seperator. default: os.linesep
    lc: multiple         Give the line color.
    bg: multiple         Give the background color.
    color_mode: str      Specify color input mode; 'names' (default), 'byte' or 'rgb'
                         see plotille.color.__docs__
Returns:
    str: histogram over `X` from left to right.

In [9]: counts = [1945, 0, 0, 0, 0, 0, 10555, 798, 0, 28351, 0]
In [10]: bins = [float('-inf'), 10, 50, 100, 200, 300, 500, 800, 1000, 2000, 10000, float('+inf')]
In [11]: print(plotille.hist_aggregated(counts, bins))

Keep in mind that there must always be n+1 bins (n is a total number of count values, 11 in the example above).

In this example the first bin is from [-inf, 10) with a count of 1945 and the last bin is from [10000, +inf] with a count of 0.

Example hist

Histogram

There is also another more 'usual' histogram function available:

In [10]: plotille.histogram?
Signature:
plotille.histogram(
    X,
    bins=160,
    width=80,
    height=40,
    X_label='X',
    Y_label='Counts',
    linesep='\n',
    x_min=None,
    x_max=None,
    y_min=None,
    y_max=None,
    lc=None,
    bg=None,
    color_mode='names',
)
Docstring:
Create histogram over `X`

In contrast to `hist`, this is the more `usual` histogram from bottom
to up. The X-axis represents the values in `X` and the Y-axis is the
corresponding frequency.

Parameters:
    X: List[float]       The items to count over.
    bins: int            The number of bins to put X entries in (columns).
    height: int          The number of characters for the height (rows).
    X_label: str         Label for X-axis.
    Y_label: str         Label for Y-axis. max 8 characters.
    linesep: str         The requested line seperator. default: os.linesep
    x_min, x_max: float  Limits for the displayed X values.
    y_min, y_max: float  Limits for the displayed Y values.
    lc: multiple         Give the line color.
    bg: multiple         Give the background color.
    color_mode: str      Specify color input mode; 'names' (default), 'byte' or 'rgb'
                         see plotille.color.__docs__

Returns:
    str: histogram over `X`.

In [11]: print(plotille.histogram(np.random.normal(size=10000)))

Example histogram

Canvas

The underlying plotting area is modeled as the Canvas class:

In [12]:  plotille.Canvas?
Init signature:
plotille.Canvas(
    width,
    height,
    xmin=0,
    ymin=0,
    xmax=1,
    ymax=1,
    background=None,
    **color_kwargs,
)
Docstring:
A canvas object for plotting braille dots

A Canvas object has a `width` x `height` characters large canvas, in which it
can plot indivitual braille point, lines out of braille points, rectangles,...
Since a full braille character has 2 x 4 dots (⣿), the canvas has `width` * 2, `height` * 4
dots to plot into in total.

It maintains two coordinate systems: a reference system with the limits (xmin, ymin)
in the lower left corner to (xmax, ymax) in the upper right corner is transformed
into the canvas discrete, i.e. dots, coordinate system (0, 0) to (`width` * 2, `height` * 4).
It does so transparently to clients of the Canvas, i.e. all plotting functions
only accept coordinates in the reference system. If the coordinates are outside
the reference system, they are not plotted.
Init docstring:
Initiate a Canvas object

Parameters:
    width: int            The number of characters for the width (columns) of the canvas.
    hight: int            The number of characters for the hight (rows) of the canvas.
    xmin, ymin: float     Lower left corner of reference system.
    xmax, ymax: float     Upper right corner of reference system.
    background: multiple  Background color of the canvas.
    **color_kwargs:       More arguments to the color-function. See `plotille.color()`.

Returns:
    Canvas object

The most interesting functions are:

point:

In [11]: plotille.Canvas.point?
Signature: plotille.Canvas.point(self, x, y, set_=True, color=None, marker=None)
Docstring:
Put a point into the canvas at (x, y) [reference coordinate system]

Parameters:
    x: float         x-coordinate on reference system.
    y: float         y-coordinate on reference system.
    set_: bool       Whether to plot or remove the point.
    color: multiple  Color of the point.
    marker: str      Instead of braille dots set a marker char.

line:

In [14]: plotille.Canvas.line?
Signature: plotille.Canvas.line(self, x0, y0, x1, y1, set_=True, color=None)
Docstring:
Plot line between point (x0, y0) and (x1, y1) [reference coordinate system].

Parameters:
    x0, y0: float    Point 0
    x1, y1: float    Point 1
    set_: bool       Whether to plot or remove the line.
    color: multiple  Color of the line.

rect:

In [15]: plotille.Canvas.rect?
Signature: plotille.Canvas.rect(self, xmin, ymin, xmax, ymax, set_=True, color=None)
Docstring:
Plot rectangle with bbox (xmin, ymin) and (xmax, ymax) [reference coordinate system].

Parameters:
    xmin, ymin: float  Lower left corner of rectangle.
    xmax, ymax: float  Upper right corner of rectangle.
    set_: bool         Whether to plot or remove the rect.
    color: multiple    Color of the rect.

text:

In [16]: plotille.Canvas.text?
Signature: plotille.Canvas.text(self, x, y, text, set_=True, color=None)
Docstring:
Put some text into the canvas at (x, y) [reference coordinate system]

Parameters:
    x: float         x-coordinate on reference system.
    y: float         y-coordinate on reference system.
    set_: bool       Whether to set the text or clear the characters.
    text: str        The text to add.
    color: multiple  Color of the point.

braille_image:

In [17]: plotille.Canvas.braille_image?
Signature:
plotille.Canvas.braille_image(
    self,
    pixels,
    threshold=127,
    inverse=False,
    set_=True,
)
Docstring:
Print an image using braille dots into the canvas.

The pixels and braille dots in the canvas are a 1-to-1 mapping, hence
a 80 x 80 pixel image will need a 40 x 20 canvas.

Example:
    from PIL import Image
    import plotille as plt

    img = Image.open("/path/to/image")
    img = img.convert('L')
    img = img.resize((80, 80))
    cvs = plt.Canvas(40, 20)
    cvs.braille_image(img.getdata(), 125)
    print(cvs.plot())

Parameters:
    pixels: list[number]  All pixels of the image in one list.
    threshold: float      All pixels above this threshold will be
                          drawn.
    inverse: bool         Whether to invert the image.
    set_: bool            Whether to plot or remove the dots.

image:

In [18]: plotille.Canvas.image?
Signature: plotille.Canvas.image(self, pixels, set_=True)
Docstring:
Print an image using background colors into the canvas.

The pixels of the image and the characters in the canvas are a
1-to-1 mapping, hence a 80 x 80 image will need a 80 x 80 canvas.

Example:
    from PIL import Image
    import plotille as plt

    img = Image.open("/path/to/image")
    img = img.convert('RGB')
    img = img.resize((40, 40))
    cvs = plt.Canvas(40, 40, mode='rgb')
    cvs.image(img.getdata())
    print(cvs.plot())

Parameters:
    pixels: list[(R,G,B)]  All pixels of the image in one list.
    set_: bool             Whether to plot or remove the background
                           colors.

plot:

In [16]: plotille.Canvas.plot?
Signature: plotille.Canvas.plot(self, linesep='\n')
Docstring:
Transform canvas into `print`-able string

Parameters:
    linesep: str  The requested line seperator. default: os.linesep

Returns:
    unicode: The canvas as a string.

You can use it for example to plot a house in the terminal:

In [17]: c = Canvas(width=40, height=20)
In [18]: c.rect(0.1, 0.1, 0.6, 0.6)
In [19]: c.line(0.1, 0.1, 0.6, 0.6)
In [20]: c.line(0.1, 0.6, 0.6, 0.1)
In [21]: c.line(0.1, 0.6, 0.35, 0.8)
In [22]: c.line(0.35, 0.8, 0.6, 0.6)
In [23]: print(c.plot())

House

Or you could render images with braille dots:

In [24]: img = Image.open('https://github.com/tammoippen/plotille/raw/master/imgs/ich.jpg')
In [25]: img = img.convert('L')
In [26]: img = img.resize((80, 80))
In [27]: cvs = Canvas(40, 20)
In [28]: cvs.braille_image(img.getdata())
In [29]: print(cvs.plot())

Me with dots

Or you could render images with the background color of characters:

In [24]: img = Image.open('https://github.com/tammoippen/plotille/raw/master/imgs/ich.jpg')
In [25]: img = img.convert('RGB')
In [25]: img = img.resize((80, 40))
In [27]: cvs = Canvas(80, 40, mode="rgb")
In [28]: cvs.image(img.getdata())
In [29]: print(cvs.plot())

Me with chars

Stargazers over time

Stargazers over time