Plot save function about roughviz HOT 6 CLOSED

@uniele @SgtChrome I have a simple solution for this (which struck me after many days of trial-and-error with different methods) which I will finish implementing this weekend.
@axmonti Yes, I will add that functionality, kindly let me know if you need any more features.

from roughviz.

The feature enhancement has been added in the latest version: V4.7.0

@uniele @SgtChrome Now you can pass the following parameter, which will render the SVG image (which then can be easily saved/copied):
plot_svg = True

@axmonti I have added the interactive parameter option as well.

I will close this issue but for any new feature enhancement, kindly open a new issue.

from roughviz.

So far I have tried a couple of ways to save the plots as png.

Consider this example, which creates html code that I saved to test.html.

import roughviz
import pandas as pd
from IPython.display import display, HTML

d = {'col1': [1, 2, 3, 4, 5], 'col2': [3, 4, 10, 2, 0.9]}
df2 = pd.DataFrame(data=d)

data = roughviz.bar(df2["col1"].values, df2["col2"].values)
display(HTML(data.data).data)

If you just wanted to render one or two images, this REST interface does them in high quality, but it's proprietary: https://htmlcsstoimage.com/demo

A solution using wkhtmltoimage failed for me and I wasn't able to figure out why, there might have been some issues with the usage of require...?

import imgkit

config = imgkit.config(wkhtmltoimage='path\to\wkhtmltoimage.exe')
options = {
    'format': 'png',
    'encoding': "UTF-8",
}
inputfile = 'path\to\test.html'
outputfile = 'path\to\out.png'
imgkit.from_file(inputfile, outputfile, config=config, options=options)

This solution using pyppeteer, which is a python wrapper for puppeteer, worked alright but produced images of lesser quality and scaling them up messed up the formatting of the plot.

import asyncio
from pyppeteer import launch

async def main():
    browser = await launch(headless=True)
    page = await browser.newPage()
    await page.setViewport({'width': 700, 'height': 400})
    await page.goto('path\to\test.html')
    await page.screenshot({'path': 'path\to\example.png'})
    await browser.close()

asyncio.get_event_loop().run_until_complete(main())

I would be very interested in what @hannansatopay were to use.

from roughviz.

Ditto for the awesome work; this is great for capturing people's attention vs. the 'usual' graphs that by now have people glossing over.

This may be a new feature or an add-on to this request, but it would be great if the 'interactive' parameter was passed along in the Python wrapper as well. The issue I'm having is that when the graph is saved as an svg (to, say, insert into PowerPoint), the highlight mask is overlaid and grouped as a large black rectangle that has to be manually removed. I've resorted to creating HTML with the interactive flag set to false and rendering that, but it would be far easier via the Python interface directly.

from roughviz.

@axmonti Would you be willing to share your way of exporting the graph to svg?

from roughviz.

Sure, but will have to 're-create' the script on my personal time, but it's pretty basic...

It's essentially a Python program that iterates through a dataset to create bar graphs for monthly KRI metrics. There are about 25 graphs, so I created a new version of the program that creates a static page on our webserver with the 25 graphs, then run PhantomJS script that iterates through svg images on the page and exports them to individual files to drag into PowerPoint (the svg file is simply the body of each svg element in the rendered webpage). My preference would be to create and export the svg files from Python, or (even better) directly from a native plot save function (as per the initial ask). I initially tried to create the whole script in PhantomJS, but ran into a few scope snags (and ran out of time).

from roughviz.