Is there the equivalent of to_markdown to read data?

Question

With pandas 1.0.0 the use of .to_markdown() to show the content of a dataframe in this forum in markdown is going to proliferate. Is there a convenient way to load the data back into a dataframe? Maybe an option to .from_clipboard(markdown=True)?

Answer 1

You can read markdown tables (or any structured text table) with the pandas read_table function:

Let's create a sample markdown table:

pd.DataFrame({"a": [0, 1], "b":[2, 3]}).to_markdown()                                                                                                                                                    

|    |   a |   b |
|---:|----:|----:|
|  0 |   0 |   2 |
|  1 |   1 |   3 |

As you can see, this is just a structured text table where the delimiters are pipes, there's a lot of whitespace, there are null columns on the left-most and right-most, and there's a header underline that must be dropped.

pd
  # Read a markdown file, getting the header from the first row and inex from the second column
  .read_table('df.md', sep="|", header=0, index_col=1, skipinitialspace=True)
  # Drop the left-most and right-most null columns 
  .dropna(axis=1, how='all')
  # Drop the header underline row
  .iloc[1:]   

   a  b
0  0  2
1  1  3