Fellow julia users,

I'd like to share a new feature in the TermWin 
<https://github.com/tonyhffong/TermWin.jl> package for navigating 
dataframes.

Basic usage:
using TermWin
using DataFrames
df = ... # read/create your dataframe
tshow( df )

With a bit of customization, you can generate very elaborate pivots to help 
get a feel of the data.

Here's an example using RDatasets's Ecdat/caschool data (the code is in 
test/dataframe.jl):

<https://lh5.googleusercontent.com/-4hvC7oq4dZE/VIrVN6BfcII/AAAAAAAAAXY/8XT9eJLTKOU/s1600/caschool.png>
Here the "top5district" pivot (after County) is a "calculated pivot". It is 
not a static column but a dynamic one that is generated based on the path 
of the tree nodes, so
as you move this pivot further up or down the pivot chain, it would be 
adjusted to the context/subdataframe correctly. Another example of 
calculated pivot is discretization on aggregated values, which can be found 
on the view selector ('v' keyboard shortcut) in the same example. You can 
also change the pivots directly using the 'p' shortcut.

The framework borrows ideas heavily from DataFramesMeta.jl (and relies on 
some of its code, too) to structure a lazy tree while keeping aggregation 
and pivoting quite performant by lifting the specifications into compiled 
functions. Numbers formatting leverages the rather performant code from 
Formatting.jl

Give it a spin and let me know how it goes.

Tony

Reply via email to