In [1]:
%run prelude.ipy
In [14]:
ax = plot.misc.total_grades_distribution(responses)
ax.figure
Out[14]:
In [47]:
axes = plot.misc.grades_by_base(responses, figsize=(20, 10))
axes[0].figure
Out[47]:
In [81]:
other_cols = ["py_years", "prog_years", "age", "degree_num", "gender_num"]
agg_cols = { "grade_value" : np.sum }
agg_cols.update({ c : np.mean for c in other_cols })
In [82]:
f = responses.groupby("exp_id").agg(agg_cols)
importances = classify.feature_importances(f, other_cols, "grade_value")
plot.misc.feature_importances(importances).figure
Out[82]:
In [83]:
ax = plot.misc.correlation_scatter(f, "prog_years", "grade_value")
ax.figure
Out[83]:
In [84]:
ax = plot.misc.correlation_scatter(f, "age", "grade_value")
ax.figure
Out[84]:
