Ch. 10 (clustering) bug

[darribas]

Cell 40 (last one of code):

ami_scores = []
# for each cluster solution
for i_cluster_type in ("k5cls", "ward5", "ward5wq", "ward5wknn"):
    # for every other clustering
    for j_cluster_type in ("k5cls", "ward5", "ward5wq", "ward5wknn"):
        # compute the adjusted mutual info between the two
        ami_score = metrics.adjusted_mutual_info_score(
            db[i_cluster_type], db[j_cluster_type]
        )
        # and save the pair of cluster types with the score
        ami_scores.append((i_cluster_type, j_cluster_type, ami_score))
# arrange the results into a dataframe
results = pandas.DataFrame(
    ami_scores, columns=["source", "target", "similarity"]
)
# and spread the dataframe out into a square
results.pivot("source", "target", "similarity")

Currently returns (with pandas 2.1.1) the following error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[40], line 17
     13 results = pandas.DataFrame(
     14     ami_scores, columns=["source", "target", "similarity"]
     15 )
     16 # and spread the dataframe out into a square
---> 17 results.pivot("source", "target", "similarity")

TypeError: pivot() takes 1 positional argument but 4 were given

[dfolch]

The following update should be backwards compatible to earlier versions of Pandas since the parameter names have not changed.

Original

results.pivot("source", "target", "similarity")

Update

results.pivot(columns="source", index="target", values="similarity")