Overview of Changes

Files added:

• setup.py - instructs python how to install the script as a package
• LICENSE.txt - includes contents of MIT license
  • replace the contents with whatever license you want
• manifest.in - instructs python to include the license file in the package
• termigraph/init.py - necessary for building package

File changed:

• moved termgraph.py to termgraph/termgraph.py - necessary for building package
• in termgraph.py - adjusted main() so it is directly callable

preperation

Review setup.py and change and adjust any value you wish, particularly:

• version
• your name
• possibly add an author_email field
• license
• description
• keywords

Review LICENSE.txt

• Make sure its the license type you want
• change your name after the copyright statement in line 3

create distribution files

• create an account on PyPi
• update setuptools, wheel and twine
  • pip3 install -U setuptools (and so likewise for wheel & twine)
• build the source distribution: python3 setup.py sdist
• build the wheel for python 3: python3 setup.py bdist_wheel

local test before submission

You now have build files, but before submitting them, you can perform a local test install

• cd into the "dist" folder and run: pip install termgraph-0.1.0-py3-none-any.whl
• this will install termgraph on your computer using the wheel generated by setup.py and link it to your path.
  • run this version by typing "termgraph" (no .py needed)
• Once you've verified that it works, you can uninstall it by typing "pip3 uninstall termgraph"

submitting to PyPi

Once you're ready to submit it to PyPi, and you've changed the values in setup.py

• build the source distribution: python3 setup.py sdist
• build the wheel for python 3: python3 setup.py bdist_wheel
• the first time you submit it to PyPi
  • register the source code: twine register dist/termgraph-0.1.0.tar.gz (Substitute your version #)
  • register the wheel: twine register dist/termgraph-0.1.0-py3-none-any.whl
• Upload the files to PyPi: twine upload dist/*

Installing on local machine in developer mode

Whenever you want to edit the script and see local changes immediately (without reinstalling each time), you can install the package in developer mode.

• First, make sure the package isn't already installed with pip
  • run pip3 uninstall termgraph
• cd into the directory containing setup.py
• type pip3 install -e .

Publishing new Releases to PyPi

• change the version # in setup.py
• run the commands to generated updated distribution & wheel files
  • the version number in the files will change
• upload the new versions
  • make sure to delete the old tar & wheel files from the "dist" directory

This is very cool, and it can me made even more useful by adding it to PyPi.

This would allow people to easily install it with pip, instead of requiring manual steps to: install pre-requisites, download the script and copy it to a location on the path.

I can submit a PR that prepares the repo for PyPi:

• adjust the script to run as a module
• create setup.py
• ~convert readme.md to readme.rst~
  • ~PyPi uses restructured text to create module documentation (not markdown)~ no longer required per @jChapman's comment below

After merging the PR, all you would need to do is:

• add a license
• create an account on PyPi
• create & upload distribution files to PyPi

Let me know if you're interested in this. As an example, here's one of my repos that's registered in PyPi

Sometimes the data set contains zero-valued data points. Since they are exactly zero, they should IMO be treated the same as when a data point is missing. However, they are currently displayed as non-zero values from the bottom quartile.

This PR changes this behaviour to treat them the same as missing data points (printing a space).

I cant initialize the termgraph command. Neither for 2.7 or 3.5:

I installed via pip and pip3

python 2.7:

[email protected]:pts/0->/home/odit (0) 
> termgraph
Traceback (most recent call last):
  File "/home/linuxbrew/.linuxbrew/bin/termgraph", line 7, in <module>
    from termgraph.termgraph import main
  File "/home/linuxbrew/.linuxbrew/Cellar/[email protected]/2.7.14_4/lib/python2.7/site-packages/termgraph/termgraph.py", line 172
    sys.stdout.write( f'\033[{color}m' ) # Start to write colorized.
                                     ^
SyntaxError: invalid syntax

Python3.5:

Traceback (most recent call last):
  File "/home/odit/.local/bin/termgraph", line 7, in <module>
    from termgraph.termgraph import main
  File "/home/odit/.local/lib/python3.5/site-packages/termgraph/termgraph.py", line 172
    sys.stdout.write( f'\033[{color}m' ) # Start to write colorized.
                                     ^
SyntaxError: invalid syntax

At the moment, termgraph uses many functions inside one script to achieve its goal. It works well, but it might be better to use classes and a more object-oriented approach. This would make it more organized, easier to build upon and easier to re-use in other projects.

The problem is that it will change a lot of the existing code structure, meaning it will break other projects which use termgraph unless they adopt these changes.

The upside is that it will make future development easier, and it will be easier for others to use this library in their own project. For example, it would make #27 easy to fix.

However, I understand if this project has progressed too far for this to be done, and restructuring all the code might not be worth the effort. But I am willing to do this myself, I have enough time available and this would be a good learning opportunity.

some error

  File "termgraph.py", line 163
    sys.stdout.write( f'\033[{color}m' ) # Start to write colorized.
                                     ^
SyntaxError: invalid syntax

i fix it by remove all "f"

sys.stdout.write( '\033[{color}m' )

The new API module as defined in #27 and #75 cannot be imported:

>>> from termgraph.module import Data, BarChart, Args, Colors
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/user/.local/share/virtualenvs/foo-IGGZy7fP/lib/python3.9/site-packages/termgraph/module.py", line 12, in <module>
    from utils import cvt_to_readable
ModuleNotFoundError: No module named 'utils'

I have made termgraph an API/module to use in other scripts and would like it to be merged with the base repository.

I have kept the structure very simple and anyone can easily extend upon it and add other types of charts to the module. Some other features have also been added.

I have made some changes in termgraph.py itself and the API/module code can be found in module.py (I know I am really bad at naming things)

Issues Handled: #27, #39, #47

Features Added:

• New param space-between that lets you print an extra space between two rows so that it looks cleaner
• Human Readable Format for values as requested (https://github.com/mkaz/termgraph/issues/43#issuecomment-580855325)

API/Module:

Colors

It houses all the colours that can be used in termgraph

Args

The Args class is basically the same thing as passing args through the command line. You can pass args as keyword arguments or use update_args on an Args instance. Args are directly passed into Chart.

Data

The Data class merges the data, labels and categories all in one. I wasn't able to add normal_data to it as well cause the normalize function required 'width' parameter for which we needed the access to the args but they are only accessible by Chart. So the only resolution was passing the args to data as well which didn't seem like a good idea. So took the normalize function and put it in Chart class. You can also visualize the data by just printing it or by using the str() function on it. It also has a property called dims (I was inspired by NumPy for this) which may or may not be helpful but I added it for future use.

Chart

The Chart class takes in the data and the args. It is drawn by using the draw() method. It is an abstract-ish class and has nothing special in it except for the _normalize() method which is the same as before and _print_header() method which prints the title and category names.

Horizontal Chart

It is derived off of the Chart class and the main reason for deriving the new chart off of this is that it contains the print_row function. I did so because no other chart required access to the function except for the ones which are horizontal.

I was only able to implement Bar Chart due to time limitations and my exams approaching and some other reasons. Below is the example code for usage (it can also be found in example.py)

from module import Data, BarChart, Args, Colors

data = Data(data=[[765, 787], [781, 769]], labels=["6th G", "7th G"], categories=["Boys", "Girls"])
chart = BarChart(data, Args(title="Total Marks Per Class", colors=[Colors.Red, Colors.Magenta], space_between=True))

chart.draw()

The display in column doesn't work properly with the --stacked option.

Example1: ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 22.00
Longertext example : ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 36.00

I have to use column -t -s ':' to clean it up.

https://www.python.org/dev/peps/pep-0563/#enabling-the-future-behavior-in-python-3-7

README.md states that the minimum version is Python 3.5+, but that is incorrect.

I intended to submit a pull request on bug fixes, however, the existing tests are largely broken.

I just clone the main branch of termgraph (to be exact, commit 63491d8b6a14dba8f88087dbdbaefb290e1ec246), and run pytest tests. It outputs a bunch of errors, showing test statistics at the end:

================ 16 failed, 21 passed in 0.63s ==================

Could you please fix these tests? @mkaz

termgraph v0.5.3

Tested on linux (and macos using gdate instead of date) with this reproduction: for X in {30..1} ; do gdate -d "-$X days" "+%F $(( $X % 10 )) " ; done | termgraph

whenever i use subprocess.call(['termgraph', 'stats.dat']) or other similar commands in a python script, i always get this error:

Traceback (most recent call last):
  File "/home/user/.local/bin/termgraph", line 8, in <module>
    sys.exit(main())
  File "/home/user/.local/lib/python3.10/site-packages/termgraph/termgraph.py", line 133, in main
    _, labels, data, colors = read_data(args)
  File "/home/user/.local/lib/python3.10/site-packages/termgraph/termgraph.py", line 712, in read_data
    colors = check_data(labels, data, args)
  File "/home/user/.local/lib/python3.10/site-packages/termgraph/termgraph.py", line 569, in check_data
    len_categories = len(data[0])
IndexError: list index out of range

however, these commands work outside python scripts like in the python shell or just the command line. is there any way i can allow them to run in python scripts as well?

I noticed that you only allow the bright colors, which are ANSI codes 90 to 96; In the AVAILABLE_COLORS enum in your main file, if additional entries were added with the codes 30 to 36, and their respective keys given something like "dark_" as a prefix, you could double the number of colors offered to users. I might try a pull request if someone else hasn't already, doesn't seem like too big a change.

Fixing the below exception when the value 0.0 is plotted

  File "/home/cmuck/Documents/Isas-script/venv/lib/python3.8/site-packages/termgraph/module.py", line 327, in print_row
    sys.stdout.write(f"\033[{Colors.black}m")  # dark gray
AttributeError: type object 'Colors' has no attribute 'black' 

Is there any ways to disable "cvt_to_readable": Return the number in a human readable format

It produces wrong numbers,

    args = {
        "stacked": False,
        "width": 50,
        "no_labels": False,
        "format": "{:>5.2f}",
        "suffix": " minutes",
        "vertical": False,
        "histogram": False,
        "no_values": False,
        "different_scale": False
    }

    chart(colors=[TERM_GRAPH_AVAILABLE_COLORS['blue']],
          data=data,
          args=args,
          labels=labels)

where data is:

[ [0.42475755214691163], [0.1093926469484965], [0.04674157698949178], [0.031145751476287842], ]