R implementation is not consistent with Python version?

[jolespin]

Am I doing this incorrectly? It doesn't seem that the clustering is computed in the same way.

from scipy.cluster.hierarchy import linkage
import pandas as pd
import numpy as np
from sklearn.datasets import load_iris

def get_iris_data():
    iris = load_iris()
    # Iris dataset
    X = pd.DataFrame(iris.data,
                     index = [*map(lambda x:f"iris_{x}", range(150))],
                     columns = [*map(lambda x: x.split(" (cm)")[0].replace(" ","_"), iris.feature_names)])

    y = pd.Series(iris.target,
                           index = X.index,
                           name = "Species")
    return X, y

# Get data
X, y = get_iris_data()

# Create an adjacency network
df_adj = np.abs(X.T.corr())

print("Maximum value: {}".format(df_adj.values.ravel().max()), "Minimum value: {}".format(df_adj.values.ravel().min()), sep="\n")

# Distance matrix
df_dism = 1 - df_adj

# Linkage 
Z = linkage(df_dism.values, method="ward", optimal_ordering=False)

# ==============
# Python version
# ==============
from dynamicTreeCut import cutreeHybrid

# Clustering
clustering_results = cutreeHybrid(Z, df_dism.values, minClusterSize=20, deepSplit=1)

Se_treecut_from_python = pd.Series(clustering_results["labels"], index=df_dism.index)
Se_treecut_from_python.head()
# Maximum value: 1.0
# Minimum value: 0.35739643082771205
# ..cutHeight not given, setting it to 30.477248373683015  ===>  99% of the (truncated) height range in dendro.
# ..done.
# iris_0    1
# iris_1    1
# iris_2    1
# iris_3    1
# iris_4    1
# dtype: int64

# ==============
# R version
# ==============
from rpy2 import robjects, rinterface
from rpy2.robjects.packages import importr
from rpy2.rinterface import RRuntimeError
from rpy2.robjects import pandas2ri
pandas2ri.activate()


r_dism = pandas2ri.py2ri(df_dism)

fastcluster = importr("fastcluster")
dynamicTreeCut = importr("dynamicTreeCut")


Z = fastcluster.hclust(R["as.dist"](r_dism), method="ward.D2")
treecut_output = dynamicTreeCut.cutreeDynamic(dendro=Z, method="hybrid", distM=r_dism, minClusterSize = 20, deepSplit=1)
Se_treecut_from_R = pd.Series(pandas2ri.ri2py(treecut_output), index=df_dism.index).astype(int)

Se_treecut_from_R.head()
# iris_0    2
# iris_1    2
# iris_2    2
# iris_3    2
# iris_4    2
# dtype: int64

[jolespin]

Any insight on this or is this project dead?

[kylessmith]

I pushed the most recent github updates to pypi. Let me know if you are still having this issue

[jolespin]

Awesome, I'd love to remove my dependency on this R package if possible.

Have you tried out r version and Python version to see if it's different with your new edits?

[jolespin]

Any updates on this issue by any chance?

[jolespin]

Just checking in about this again.