Save to remote file system

[ajmeese7]

Hey there, quick question that I am trying to help someone with. They are looking to save their dif-generated image directly to an S3 bucket on AWS without saving the image locally. Do you know of any ways to specify this currently with the library?

[PaleNeutron]

I am not familiar with S3 bucket but I think it will provide some api to upload files.

You could save file to disk or BytesIO and then upload it by python code.

[ajmeese7]

@PaleNeutron do you have an example of how to convert the image generated by this library to BytesIO without saving it to the disk? That would be greatly appreciated!

[PaleNeutron]

import pandas as pd
import numpy as np
import dataframe_image as dfi
from io import BytesIO

df = pd.DataFrame(np.random.rand(5,8))

df

	0	1	2	3	4	5	6	7
0	0.018129	0.511246	0.210085	0.174160	0.562191	0.533760	0.836536	0.058780
1	0.452144	0.718135	0.350870	0.881881	0.708918	0.990391	0.864563	0.440264
2	0.154236	0.069712	0.580959	0.170795	0.621348	0.202536	0.964883	0.566463
3	0.373485	0.480376	0.956419	0.230610	0.386035	0.750750	0.437797	0.030412
4	0.001774	0.984068	0.421984	0.026590	0.665637	0.395694	0.141108	0.537476

png_io = BytesIO()
df.dfi.export(png_io)

data = png_io.getbuffer()

len(data)

33341

type(data)

memoryview

[PaleNeutron]

You can get bytes from data, and send it to any where you want

[ajmeese7]

You're the best, thanks for your help!