emoen / machine-learning-for-asset-managers Goto Github PK

For the cMDA function in ch6_feature_importance_analysis.py, should the following code be outside of train test for loop? The mean score is for cross vaildation calcualted on cluster level.

    imp=(-1*scr1).add(scr0,axis=0)
    imp = imp/(-1*scr1)
    imp = pd.concat({'mean':imp.mean(), 'std':imp.std()*imp.shape[0]**-.5}, axis=1)
    imp.index=['C_'+str(i) for i in imp.index]

There are some parts that do not work due to typos or name errors when the module is executed as a script.

ch3_metrics.py: from l38

- def mutualInfor(x,y, norm=False):
+ def mutualInfo(x,y, norm=False):
    #mutual information
    bXY = numBins(x.shape[0], corr = np.corrcoef(x,y)[0,1])
    cXY = np.histogram2d(x,y, bXY)[0]
    iXY = mutual_info_score(None, None, contingency=cXY)
    if norm:
-        hX = ss.entropy(np.histogram(x, bins)[0]) #marginal 
-        hY = ss.entropy(np.histogram(y, bins)[0]) #marginal
+        hX = ss.entropy(np.histogram(x, bXY)[0]) #marginal 
+        hY = ss.entropy(np.histogram(y, bXY)[0]) #marginal
        iXY /= min(hX, hY) #normalized mutual information

ch7_portfolio_construction.py: from l120

    # code snippet 7.4 - intracluster optimal allocations
    # step 2. compute intracluster allocations using the denoised cov matrix
    wIntra = pd.DataFrame(0, index=cov0.index, columns=clstrs.keys())
    for i in clstrs:
-        wIntra.loc[clstrs[i], i] = minVarPort(cov1.loc[clstrs[i], clstrs[i]]).flatten()
+        wIntra.loc[clstrs[i], i] = minVarPort(corr1.loc[clstrs[i], clstrs[i]]).flatten()

When used as a package, the description from ch4_optimal_clustering import clusterKMeansBase in l13 of "ch6_feature_importance_analysis.py" causes the error "ModuleNotFoundError: No module named 'ch4_optimal_clustering'". The reference needs to be changed to an explicit relative reference from .ch4_optimal_clustering import clusterKMeansBase.

The description fit.enumerators_ in l33 of "ch6_feature_importance_analysis.py" does not work correctly because the attribute name is wrong, causing the error "AttributeError: 'BaggingClassifier' object has no attribute 'enumerators_'". The correct attribute is fit.estimators_.

When I run the script called 'ch7_portfolio_construction' I get the following error:

AttributeError Traceback (most recent call last)
in
39 cov1 = pd.DataFrame(cov1, index=cols, columns=cols)
40 corr1 = mp.cov2corr(cov1)
---> 41 corr1, clstrs, silh = oc.clusterKMeansBase(pd.DataFrame(corr0))
42
43 # # code snippet 7.4 - intracluster optimal allocations

D:\User\Desktop\ch4_optimal_clustering.py in clusterKMeansBase(corr0, maxNumClusters, n_init, debug)
56 print("********")
57
---> 58 newIdx = np.argsort(kmeans.labels_)
59 #print(newIdx)
60

AttributeError: 'NoneType' object has no attribute 'labels_'

Can you please help me to fix this bug? Thank you!