Search code examples
pythonvisualizationdata-visualizationaspect-ratioimshow

How to render visually a big pandas dataframe (~7000 lines, 3 columns)?

I have a big dataframe with values between -1 and 1. The dataframe has almost 7000 lines and 3 columns. I would like to have a visualization without the cells shrinking. For now, it's more lines than cells. I would like the color to refer to the value. I was trying with imshow but I don't find how to increase the height of cells. I've tried to play with extent and aspect but when aspect is increased, the picture shrinks.

Solution

  • The most compact representation would be to use a bitmap of one pixel per value. This can be accomplished using PIL or Pillow, by creating the image from a buffer of data. 

    import StringIO
from PIL import Image

im = Image.open(StringIO.StringIO(buffer))

    You would have to scale the data in the buffer to one or more bytes per pixel depending on the mode of the image you want to create