epmoyer / ipy_table Goto Github PK

Or option to drop "|" between two columns for specific rows.

A semicolon is missing in:

  text = text.replace(' ', '&nbsp')

The correct replacement for a non-breaking space is  

Hi,

I think you probably want to make a link from
http://nbviewer.ipython.org/urls/raw.github.com/epmoyer/ipy_table/master/ipy_table-Introduction.ipynb
to http://epmoyer.github.com/ipy_table/
it would be easier to navigate .

Hi

I was wondering if there is a way to use your package to display several data frames side by side in a jupyter notebook cell. kind of being able to replicate the powepoint slides from this youtube video

https://www.youtube.com/watch?v=9d5-Ti6onew

thanks in advance
Dragos

Is it possible to make a cell or a row a hyperlink to another URL? I tried using IPython.display.HTML() but to no effect.

As best I can tell, it amounts to changing the single print statement to the print function, with an appropriate import from __future__, and swapping out xrange for range.

Or perhaps simply enabling the 2to3 tool in setup.py

For example, if I have a list of Foo, and repr(foo) is <Foo at 0x49199d0>, that's what I should see in the table cell. Right now the cell gets that literally and it's interpreted as an XML tag, which isn't so great.

Objects that have specific HTML repr methods (_repr_html_ for ipython-notebook) should use those instead of being escaped.

MarkupSafe may also be of interest here.

Would be very helpful

From what I can see, ipy_table works best if it is the last row in a ipython notebook cell. If not, then there are some issues. I can only get a formatted table if make_table and apply_theme is the last statements.

More info:
I have a loop that will create plots. After each plot I want to have a table. The only way I got ipy_table to output anything was using:
display(make_table(table_content))

The problem is that I could not get a formatted table this way.

If my table starts with a header row, I can use apply_theme('basic') to highlight it.

But if my table doesn't start with a header row, and I want to display names for the columns, I'd like to be able to specify them when calling make_table()

But please make sure to bump the version first :)

If I try to create a table with strings that contain non-ascii characters, I get a UnicodeEncodeError:

ipy_table.make_table([(u'\xe9', 1)])

This produces:

----> 3 ipy_table.make_table([(u'\xe9', 1)])

/usr/lib/python2.7/site-packages/ipy_table.pyc in make_table(array, interactive, debug)
    421     global _TABLE
    422     _TABLE = IpyTable(array, interactive=interactive, debug=debug)
--> 423     return _TABLE._render_update()
    424 
    425 

/usr/lib/python2.7/site-packages/ipy_table.pyc in _render_update(self)
    233         """Renders the table only if in interactive mode."""
    234         if(self._interactive):
--> 235             return self.render()
    236         return None
    237 

/usr/lib/python2.7/site-packages/ipy_table.pyc in render(self)
    221     def render(self):
    222         """Render the table.  Return an iPython IPython.core.display object."""
--> 223         html = self.get_table_html()
    224         if self._debug:
    225             print html

/usr/lib/python2.7/site-packages/ipy_table.pyc in get_table_html(self)
    181                     # (if it is a float)
    182                     item_html = self._formatter(
--> 183                         item, self._cell_styles[row][column])
    184 
    185                     # Add bold and italic tags if set

/usr/lib/python2.7/site-packages/ipy_table.pyc in _formatter(self, item, cell_style)
    378             text = cell_style['float_format'] % item
    379         else:
--> 380             text = str(item)
    381 
    382         # If cell wrapping is not specified

UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 0: ordinal not in range(128)

Please excuse the brievety I write from my phone.

May I suggest using link to nbviewer on github pages for easy viewing.

Also instead of returning an HTML object, return a dummy object having several _repr_xxx so that each frontend can choose. Also notebook will store repr then nbconvert can use repr latex or other when converting ipynb to pdf.

We will move notebook to bootstrap which have convenience css for table (alternate row color, highlight on hoover) I think you might be interested for futur update.

I will have a deeper look later.

Did you add a link to your project on ipython wiki ?

I was surprised that a statement like

ipy_table.set_column_style(1, align='right')

would re-render a table I had worked with before. How does this work? There seems to be some sort of global state that the library is maintaining, holding on to something like the most recently-generated table.
In these days of object-oriented programming, functional programming and parallel processing I think maintaining state like that can often be problematic. What issues does it help with, and what issues does it cause?