Saturday, July 3, 2021

Plotext – Python Plotting on the Terminal

example

plotext plots directly on terminal, it has no dependencies and the syntax is very similar to matplotlib. It also provide a simple command line tool.

Note: there are many new feautures from the previous version, any bug report is usefull and very welcomed.

Table of Contents

Scatter Plot

Here is a basic example of a scatter plot:

import plotext as plt
y = plt.sin(100, 3) # sinuisodal signal with 100 points and 3 periods
plt.scatter(y)
plt.plotsize(100, 30)
plt.title("Scatter Plot Example")
plt.show()

which prints this on terminal: example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; y = plt.sin(100, 3); plt.scatter(y); plt.plotsize(100, 30); plt.title('Scatter Plot Example'); plt.show()"

Access the scatter docstring for more documentation.

Note: the higher resolution marker shown in the picture doesnt work in Windows for now, use other available markers in this case, like dot, big or others. See the section Plot Aspect for further guidance.

Table of Contents

Line Plot

For a line plot use the the plot function instead:

import plotext as plt
y = plt.sin(100, 3)
plt.plot(y)
plt.plotsize(100, 30)
plt.title("Plot Example")
plt.show()

example The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; y = plt.sin(100, 3); plt.plot(y); plt.plotsize(100, 30); plt.title('Plot Example'); plt.show()"

Access the plot docstring for more documentation.

Table of Contents

Log Plot

For a log plot use the the xscale() or yscale() functions after the plotting functions:

import plotext as plt
l = 10 ** 4
x = range(1, l + 1)
y = plt.sin(l, 2)
plt.plot(x, y)
plt.plotsize(100, 30)
plt.xscale("log")
plt.yscale("linear")
plt.title("Logarithmic Plot")
plt.xlabel("logarithmic scale")
plt.ylabel("linear scale")
plt.grid(1, 0)
plt.show()

example The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; l = 10 ** 4; x = range(1, l + 1); y = plt.sin(l, 2); plt.plot(x, y); plt.plotsize(100, 30); plt.xscale('log'); plt.yscale('linear'); plt.title('Logarithmic Plot'); plt.xlabel('logarithmic scale'); plt.ylabel('linear scale'); plt.grid(1, 0); plt.show()"

Access the xscale and yscale docstring for more documentation.

Table of Contents

Stem Plot

For a stem plot use either fillx or filly parameters. Here is a bisic example:

import plotext as plt
y = plt.sin(50, 2)
plt.scatter(y, fillx = True)
plt.plotsize(100, 30)
plt.title("Stem Plot")
plt.show()

example The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; y = plt.sin(50, 2); plt.scatter(y, fillx = True); plt.plotsize(100, 30); plt.title('Stem Plot'); plt.show()"

Table of Contents

Multiple Data Sets

Multiple data sets can be plotted using consecutive scatter or plot functions. Here is a basic example:

import plotext as plt
y1 = plt.sin(1000, 3)
y2 = plt.sin(1000, 3, 1.5, phase = 1)
plt.plot(y1, label = "plot")
plt.scatter(y2, label = "scatter", marker = "small")
plt.plotsize(100, 30)
plt.title("Multiple Data Set")
plt.show()

Using the label parameter inside the plotting calls, a legend is automatically added in the upper left corner of the plot.

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; y1 = plt.sin(1000, 3); y2 = plt.sin(1000, 3, 1.5, phase = 1); plt.plot(y1, label = 'plot'); plt.scatter(y2, label = 'scatter', marker = 'small'); plt.plotsize(100, 30); plt.title('Multiple Data Set'); plt.show()"

Table of Contents

Double Y Axis

Data could be plotted indipendently on both left and right y axes, using the yaxis parameter. Here is a simple example:

import plotext as plt
y1 = plt.sin(1000, 3)
y2 = [2 * el for el in plt.sin(1000, 1, 0, phase = 1)]
plt.plot(y1, label = "plot", yaxis = "left")
plt.scatter(y2, label = "scatter", marker = "small", yaxis = "right")
plt.plotsize(100, 30)
plt.title("Double Y Axis")
plt.ylabel("left axis", "right axis")
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; y1 = plt.sin(1000, 3); y2 = [2 * el for el in plt.sin(1000, 1, 0, phase = 1)]; plt.plot(y1, label = 'plot'); plt.scatter(y2, label = 'scatter', marker = 'small', yaxis = 'right'); plt.plotsize(100, 30); plt.title('Double Y Axis'); plt.ylabel('left axis', 'right axis'); plt.show()"

The yaxis parameter is also used in the functions yscale, yticks, ylim, plot, bar and hist.

Table of Contents

Bar Plot

For a bar plot use the the bar function. Here is an example:

import plotext as plt
cities = ["Tokyo", "Delhi", "Shanghai", "São Paulo", "Mexico City", "Cairo", "Mumbai", "Beijing"]
population = [37400068, 28514000, 25582000, 21650000, 21581000, 20076000, 19980000, 19618000]
plt.bar(cities, population)
plt.plotsize(100, 30)
plt.title("Bar Plot of the World Largest Cities")
plt.xlabel("City")
plt.ylabel("Population")
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; cities = ['Tokyo', 'Delhi', 'Shanghai', 'São Paulo', 'Mexico City', 'Cairo', 'Mumbai', 'Beijing']; population = [37400068, 28514000, 25582000, 21650000, 21581000, 20076000, 19980000, 19618000]; plt.bar(cities, population); plt.plotsize(100, 30); plt.title('Bar Plot of the World Largest Cities'); plt.xlabel('City'); plt.ylabel('Population'); plt.show()"

Access the bar docstring for more documentation. Note: for now it doesn't work with log scale.

Note: the higher resolution marker shown in the picture doesnt work in Windows for now, use other available markers in this case, like dot, big or others. See the section

Table of Contents

Histogram Plot

For a histogram plot use the the hist function. Here is an example:

import plotext as plt
import random
l = 10 ** 3
data1 = [random.gauss(0, 1) for el in range(10 * l)]
data2 = [random.gauss(3, 1) for el in range(6 * l)]
data3 = [random.gauss(6, 1) for el in range(4 * l)]
plt.clp()
bins = 60
plt.hist(data1, bins, label="mean 0")
plt.hist(data2, bins, label="mean 3")
plt.hist(data3, bins, label="mean 6")
plt.title("Histogram Plot")
plt.xlabel("data bin")
plt.ylabel("frequency")
plt.plotsize(100, 30)
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; import random; l = 10 ** 3; data1 = [random.gauss(0, 1) for el in range(10 * l)]; data2 = [random.gauss(3, 1) for el in range(6 * l)]; data3 = [random.gauss(6, 1) for el in range(4 * l)]; plt.clp(); bins = 60; plt.hist(data1, bins, label='mean 0'); plt.hist(data2, bins, label='mean 3'); plt.hist(data3, bins, label='mean 6'); plt.title('Histogram Plot'); plt.xlabel('data bin'); plt.ylabel('frequency'); plt.plotsize(100, 30); plt.show()"

Access the hist docstring for more documentation. Note: for now it doesn't work with log scale.

Table of Contents

Data Ticks

You can change the numerical ticks on both axes with the following three functions - to be placed before show():

  • plt.ticks(xnum, ynum) sets the ticks frequency on respectivelly the x and y axis.
  • plt.xticks(ticks, labels) manually sets the x ticks to the list of labels at the list of coordinates provided in ticks. If only one list is provided (ticks), the labels will correspond to the coordinates.
  • plt.yticks(ticks, labels) is the equivalent of plt.xticks() but for the y axis. It also takes the optional parameter yaxis in case multiple y axes are used in the plot.

Here is a coded example:

import plotext as plt
l, n = 1000, 3
y1 = plt.sin(l, n)
y2 = plt.sin(l, n, 2)
import numpy as np
xticks = np.arange(0, l + l / (2 * n), l / (2 * n))
xlabels = [str(i) + "π" for i in range(2 * n + 1)]
plt.scatter(y1, label = "periodic signal")
plt.scatter(y2, label = "decaying signal", marker = "small", color = "gold")
plt.plotsize(100, 30)
plt.ticks(None, 3)
plt.xticks(xticks, xlabels)
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; l, n = 1000, 3; y1 = plt.sin(l, n); y2 = plt.sin(l, n, 2); import numpy as np; xticks = np.arange(0, l + l / (2 * n), l / (2 * n)); xlabels = [str(i) + 'π' for i in range(2 * n + 1)]; plt.scatter(y1, label = 'periodic signal'); plt.scatter(y2, label = 'decaying signal', marker = 'small', color = 'gold'); plt.plotsize(100, 30); plt.ticks(None, 3); plt.xticks(xticks, xlabels); plt.show()"

Access the ticks, xticks, yticks docstrings for more documentation.

Table of Contents

Date Time Plot

To plot dates and/or times use the string_to_time() function like in the example below:

import plotext as plt
plt.clp()
dates = ["01/01/2021 12:20", "02/01/2021 15:40", "03/01/2021 15:10", "04/01/2021 15:24", "05/01/2021", "05/01/2021 18:10", "05/01/2021 23:20", "06/01/2021 12:10","07/01/2021"]
prices = [100, 110, 130, 140, 150, 160, 170, 180]
dates_x = [plt.string_to_time(el) for el in dates] 
plt.plot(dates_x, prices, marker = ".")
plt.scatter(dates_x, prices, marker = "small")
plt.xticks(dates_x, dates)
plt.title("Date-Time Plot")
plt.xlabel("Date-Time")
plt.ylabel("Stock Price $")
plt.plotsize(100, 30)
plt.show()

example The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; plt.clp(); dates = ['01/01/2021 12:20', '02/01/2021 15:40', '03/01/2021 15:10', '04/01/2021 15:24', '05/01/2021', '05/01/2021 18:10', '05/01/2021 23:20', '06/01/2021 12:10','07/01/2021']; prices = [100, 110, 130, 140, 150, 160, 170, 180]; dates_x = [plt.string_to_time(el) for el in dates]; plt.plot(dates_x, prices, marker = '.'); plt.scatter(dates_x, prices, marker = 'small'); plt.xticks(dates_x, dates); plt.title('Date-Time Plot'); plt.xlabel('Date-Time'); plt.ylabel('Stock Price $'); plt.plotsize(100, 30); plt.show()"

Access the string_to_time docstrings for more documentation.

Table of Contents

Plot Limits

The plot limits are set automatically, to set them manually you can use the following functions - to be placed before show():

  • plt.xlim(xmin, xmax) sets the minimum and maximum limits of the plot on the x axis. It requires a list of two numbers, where the first xmin sets the left (minimum) limit and the second xmax the right (maximum) limit. If one or both values are not provided, they are calculated automatically.
  • plt.ylim(ymin, ymax) is the equivalent of plt.xlim() but for the y axis. It also takes the optional parameter yaxis in case multiple y axes are used in the plot.

Here is a coded example:

import plotext as plt
l, n = 1000, 2
x = range(1, l + 1)
y = plt.sin(l, n)
plt.scatter(x, y, color = "indigo")
plt.xlim(x[0] - 101, x[-1] + 100)
plt.ylim(-1.2, 1.2)
plt.plotsize(100, 30)
plt.show()
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; l, n = 1000, 2; x = range(1, l + 1); y = plt.sin(l, n); plt.scatter(x, y, color = 'indigo'); plt.xlim(x[0] - 101, x[-1] + 100); plt.ylim(-1.2, 1.2); plt.plotsize(100, 30); plt.show()"

Table of Contents

Plot Aspect

You can personalize the plot aspect in many ways. You could use the following parameters - to be placed inside the scatter, plot, bar and hist calls:

  • marker sets the marker used to identify each data point to the specified character. For example plt.scatter(data, marker = "x"). Access the markers() function for further marker codes.
  • color = color sets the color of the marker used. Access the colors() function for the color codes available.
  • fillx = True fills the area between the data and the x axis with data points (if used inside scatter) or line points (if used inside plot). For example: plt.plot(data, fillx = True). By default fillx = False
  • filly is the correspondent parameter of fillx but for the y axis.
  • Note that the functions bar and hist use the parameter fill (instead of fillx and filly) which is used to fill the bars with colors

You could also use the following functions - to be placed before show():

  • plotsize(width, height) sets the width and height of the plot to the desired values. Note that the plot automatically extends to fill the entire terminal: use this function in order to reduce this size.
  • title(string) adds a plot title on the top of the plot.
  • xlabel(string) and ylabel(string) adds a label for respectively the x and y axis on the bottom of the plot. If two strings are provided to plt.ylabel() the second is indended for the right y axis.
  • grid(xbool, ybool) adds the x grid lines to the plot if xbool == True and the y grid lines if ybool == True. If only one Boolean value is provided both grid lines are set simultaneously.
  • xaxes(bool1, bool2) adds the lower x axis if bool1 == True and the upper x axis if bool2 == True. If only one boolean value is provided both axes are set simultaneously.
  • yaxes(bool1, bool2) adds the left y axis if bool1 == True and the right y axis if bool2 == True. If only one boolean value is provided both axes are set simultaneously.
  • canvas_color(color) sets the color of the plot canvas alone (the area where the data is plotted).
  • axes_color(color) sets the background color of all the labels surrounding the actual plot, i.e. the axes, axes labels and ticks, title and legend, if present.
  • ticks_color(color) sets the (full-ground) color of the axes ticks and of the grid lines, if present.
  • colorless() (in short cls()) removes all colors from the current plot.

Here is a coded example:

import plotext as plt
l, n = 1000, 2
x = range(1, l + 1)
y = plt.sin(l, n)
plt.plot(x, y1, label = "periodic signal", color = "violet", marker = "small")
plt.plotsize(100, 30)
plt.grid(True)
plt.title("Plot Style Example")
plt.xlabel("x axis label")
plt.ylabel("y axis label")
plt.canvas_color("white")
plt.axes_color("cloud")
plt.ticks_color("iron")
plt.xaxes(1, 0)
plt.yaxes(1, 0)
plt.ticks(10)
plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; l, n = 1000, 2; x = range(1, l + 1); y = plt.sin(l, n); plt.plot(x, y, label = 'periodic signal', color = 'violet', marker = 'small'); plt.plotsize(100, 30); plt.grid(True); plt.title('Plot Style Example'); plt.xlabel('x axis label'); plt.ylabel('y axis label'); plt.canvas_color('white'); plt.axes_color('cloud'); plt.ticks_color('iron'); plt.xaxes(1, 0); plt.yaxes(1, 0); plt.ticks(10); plt.show()"

Here are the colors and markers codes:

colors

Note: using flash will result in an actual white flashing marker (therefore it will not work with white canvas background color).

colors

Table of Contents

Multiplie Subplots

In order to plot a grid of plots, use the following main functions:

  • plt.subplots(rows, cols) creats a matrix of plots with the given number of rows and columns.
  • plt.subplot(row, col) access the plot at the given row and column, counting (from 1) from the upper left corner of the matrix of plots, previously set.

Here is a coded basic example:

import plotext as plt
l, n = 1000, 5
y = plt.sin(l, n)

plt.clf()
plt.subplots(2, 1)

plt.subplot(1, 1)
plt.plot(y, yaxis = "right")
plt.plotsize(100, 27)

plt.subplot(2, 1)
plt.hist(y, color = "indigo")
plt.plotsize(100, 27)

plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; l, n = 1000, 5; y = plt.sin(l, n); plt.clf(); plt.subplots(2, 1); plt.subplot(1, 1); plt.plot(y, yaxis = 'right'); plt.plotsize(100, 27); plt.subplot(2, 1); plt.hist(y, color = 'indigo'); plt.plotsize(100, 27); plt.show()"

Table of Contents

Streaming Data

When streaming a continuos flow of data, consider using the following functions:

  • clear_figure (in short clf) clears the entire figure, including its subplots.
  • clear_plot() (in short clp()) clears the plot and all its internal parameters; it is useful when running the same script several times in order to avoid adding the same data to the plot; it is very similar to cla() in matplotlib.
  • clear_data() (in short cld()) clear only the plot data (without clearing the plot style).
  • clear_terminal() (in short clt()) clear the terminal before the actual plot.
  • sleep(time) is used in order to reduce a possible screen flickering; for example sleep(0.01) would add approximately 10 ms to the computation. Note that the time parameters will depend on your processor speed and it needs some manual tweaking.
  • The function colorless() is recommended to make the streaming more responsive, but not mandatory.

Here is a coded example:

import plotext as plt
import numpy as np
l, n = 1000, 2
x = np.arange(0, l)
xticks = np.linspace(0, l - 1, 5)
xlabels = [str(i) + "π" for i in range(5)]
frames = 100
    
plt.clf()
plt.ylim(-1, 1)
plt.xticks(xticks, xlabels)
plt.yticks([-1, 0, 1])
plt.plotsize(100, 30)
plt.title("Streaming Data")
plt.colorless()

for i in range(frames):
    y = plt.sin(l, n, 0, phase = 2 * i  / frames)       

    plt.cld()
    plt.clt()
    plt.scatter(x, y, marker = "dot")
    plt.sleep(0.01)
    plt.show()

example

The equivalent direct terminal command line is:

python3 -m plotext "import plotext as plt; import numpy as np; l, n = 1000, 2; x = np.arange(0, l); xticks = np.linspace(0, l - 1, 5); xlabels = [str(i) + 'pi' for i in range(5)]; frames = 100; plt.clf(); plt.ylim(-1, 1); plt.xticks(xticks, xlabels); plt.yticks([-1, 0, 1]); plt.plotsize(100, 30); plt.title('Streaming Data'); plt.colorless();  y = lambda i: plt.sin(l, n, 0, phase = 2 * i  / frames); [(plt.cld(), plt.scatter(x, y(i), marker = 'dot'), plt.sleep(0.01), plt.show()) for i in range(frames)]"

Plotting the same data using matplotlib was roughly 10 to 50 times slower on my Linux-based machine (depending on the colors settings and data size).

Table of Contents

Other Functions

  • savefig(path) saves the plot as a text file at the path provided. Note: no colors are preserved at the moment, when saving.
  • get_canvas() return the plot final canvas as a string. To be used after the show function possibly with its hide parameter set to True.
  • version() returns the version of the current installed plotext package.
  • sin() returns a sinusoidal function usefull for testing. Access its docstring for durther documentation.
  • plt.docstrings() prints all the available doc-strings.
  • test() runs all the above tests in sequence:
import plotext as plt
plt.test()

Table of Contents

Installation

To install the latest version of plotext use: pip install plotext --upgrade or in Linux sudo -H pip install plotext --upgrade

Table of Contents

Main Updates:

  • from version 3.1.0: fixed issue of plot resizing reported by @nicrip
  • from version 3.0.1: added clear_data() and test() functions

from version 2:

  • Direct terminal command line tool added
  • Smaller marker added (with improved resolution), with new marker codes
  • Subplots added
  • Log plots added
  • Stem plot added
  • Double Y Axes added
  • Bar plot added
  • Date/Time Plot added
  • get_canvas() function added
  • sin() function added
  • clear_figure() function added
  • figsize() changed to plotsize()
  • nocolor() changed to colorless()
  • frame option removed and replaced with xaxes and yaxes.
  • most of the code re-written

Table of Contents

Future Plans:

Under request (just open an issue report about it):

  • higher resolution marker support for Windows (the one named small)
  • log scale for bar/hist plot
  • subplots with columnspan and rowspan parameters
  • Spider/Idle terminal support or for other more rare terminals (if possible)
  • saving plot text files with color (not sure if usefull)

Any help or new ideas are welcomed.

Table of Contents



from Hacker News https://ift.tt/3yoPs2d
Posted by at

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

Subscribe to: Post Comments (Atom)