test_R2[name].append(model.score(X[test,:],y[test]))
test_pred = model.predict(X[test,:])
if name == 'RF': #test feature importance for RF model
for col_name in use_cols:
Xshuff = X[test,:].copy()
col_idx = col_dict[col_name]
obj = slice(col_idx[0],col_idx[-1]+1)
shuff_R2 = np.zeros(n_feature_shuffs)
for n in np.arange(n_feature_shuffs):
np.random.shuffle(Xshuff[:,obj])
shuff_R2[n] = model.score(Xshuff, y[test])
RF_feature_imp[col_name].append(test_R2[name][-1] - np.mean(shuff_R2))
returns[name].append(LCM.pick_K_returns(
test_pred, net_returns[test], prnc_weights[test],
pick_K_list, n_boots=100, sub_marg=False))
grade_returns[name].append(LCM.pick_K_returns_by_grade(
test_pred, net_returns[test], prnc_weights[test],
LD.iloc[test][grade_group], grade_pick_K))
grade_makeup[name][cnt,:] = LCM.get_choice_grade_makeup(test_pred, LD.iloc[test][grade_group],
unique_grades, grade_pick_K)
cnt += 1
# Annualize portfolio returns, and convert them into numpy arrays as needed
rel_returns = {}
marg_returns = LCM.annualize_returns(np.array(marg_returns))
for name, model in model_set:
returns[name] = LCM.annualize_returns(np.array(returns[name]))
if name == 'RF': #test feature importance for RF model
for col_name in use_cols:
Xshuff = X[test, :].copy()
col_idx = col_dict[col_name]
obj = slice(col_idx[0], col_idx[-1] + 1)
shuff_R2 = np.zeros(n_feature_shuffs)
for n in np.arange(n_feature_shuffs):
np.random.shuffle(Xshuff[:, obj])
shuff_R2[n] = model.score(Xshuff, y[test])
RF_feature_imp[col_name].append(test_R2[name][-1] -
np.mean(shuff_R2))
returns[name].append(
LCM.pick_K_returns(test_pred,
net_returns[test],
prnc_weights[test],
pick_K_list,
n_boots=100,
sub_marg=False))
grade_returns[name].append(
LCM.pick_K_returns_by_grade(test_pred, net_returns[test],
prnc_weights[test],
LD.iloc[test][grade_group],
grade_pick_K))
grade_makeup[name][cnt, :] = LCM.get_choice_grade_makeup(
test_pred, LD.iloc[test][grade_group], unique_grades, grade_pick_K)
cnt += 1
# Annualize portfolio returns, and convert them into numpy arrays as needed
rel_returns = {}
