User interface to statsmodels and pandas
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.



Grimm is a program for manipulating statistic and econometric data, analogously to Gretl, STATA...

It's in a very early stage of development, but to get an idea of how it looks you can just download it and run "./grimm". Then, paste into the scripts editing window the following:

ols(PetalWidth, SepalWidth)
scatter(PetalWidth, SepalWidth)

and click the "Run" button just above.


In order to run Grimm, you will need to have installed:

  • Python,
  • gtk 3 libraries, including GtkSourceView and support for introspection (all rather standard in modern Linux distributions),
  • pandas,
  • matplotlib,
  • a recent (0.4.2 is not enough) version of statsmodels.


  • to be scriptable in python;
  • to have a nice and intuitive GUI (with a multiwindows mode like Gretl, and a single window mode as Stata among which to choose);
  • to have perfect integration between GUI, scripts and interactive shell: not all commands/options will necessarily be exposed in the GUI, but everything which is exposed in the GUI will be "scriptable", and it will be as easy as possible to do so (i.e. commands "given" through the GUI will appear in the shell's history);
  • to never reinvent the wheel, but rather build an abstract enough skeleton so as to make it as quick as possible to support new functionalities from statsmodels/pandas;
  • to allow handling multiple datasets together,
  • to ease personalization/expansion, while guaranteeing consistency,
  • to make replicability and integration in the academic workflow trivial: it should require only a minimal effort to make a script produce all the graphs, tables and number ready to be used in a LaTeX document.