GetFPType fix

Author: Alexsandruss

cuml/bench.py

@@ -10,6 +10,33 @@
 import json
 
 
+def get_dtype(data):
+    '''
+    Get type of input data as numpy.dtype
+    '''
+    if hasattr(data, 'dtype'):
+        return data.dtype
+    elif hasattr(data, 'dtypes'):
+        return str(data.dtypes[0])
+    elif hasattr(data, 'values'):
+        return data.values.dtype
+    else:
+        raise ValueError(f'Impossible to get data type of {type(data)}')
+
+
+try:
+    from daal4py.sklearn._utils import getFPType
+except ImportError:
+    def getFPType(X):
+        dtype = str(get_dtype(X))
+        if 'float32' in dtype:
+            return 'float'
+        elif 'float64' in dtype:
+            return 'double'
+        else:
+            ValueError('Unknown type')
+
+
 def sklearn_disable_finiteness_check():
     try:
         sklearn.set_config(assume_finite=True)
@@ -427,18 +454,6 @@ def convert_data(data, dtype, data_order, data_format):
         return cudf.DataFrame.from_pandas(pd.DataFrame(data))
 
 
-def get_dtype(data):
-    '''
-    Get type of input data as numpy.dtype
-    '''
-    if hasattr(data, 'dtype'):
-        return data.dtype
-    elif hasattr(data, 'dtypes'):
-        return str(data.dtypes[0])
-    elif hasattr(data, 'values'):
-        return data.values.dtype
-
-
 def read_csv(filename, params):
     from string import ascii_lowercase, ascii_uppercase
 

daal4py/bench.py

@@ -10,6 +10,33 @@
 import json
 
 
+def get_dtype(data):
+    '''
+    Get type of input data as numpy.dtype
+    '''
+    if hasattr(data, 'dtype'):
+        return data.dtype
+    elif hasattr(data, 'dtypes'):
+        return str(data.dtypes[0])
+    elif hasattr(data, 'values'):
+        return data.values.dtype
+    else:
+        raise ValueError(f'Impossible to get data type of {type(data)}')
+
+
+try:
+    from daal4py.sklearn._utils import getFPType
+except ImportError:
+    def getFPType(X):
+        dtype = str(get_dtype(X))
+        if 'float32' in dtype:
+            return 'float'
+        elif 'float64' in dtype:
+            return 'double'
+        else:
+            ValueError('Unknown type')
+
+
 def sklearn_disable_finiteness_check():
     try:
         sklearn.set_config(assume_finite=True)
@@ -427,18 +454,6 @@ def convert_data(data, dtype, data_order, data_format):
         return cudf.DataFrame.from_pandas(pd.DataFrame(data))
 
 
-def get_dtype(data):
-    '''
-    Get type of input data as numpy.dtype
-    '''
-    if hasattr(data, 'dtype'):
-        return data.dtype
-    elif hasattr(data, 'dtypes'):
-        return str(data.dtypes[0])
-    elif hasattr(data, 'values'):
-        return data.values.dtype
-
-
 def read_csv(filename, params):
     from string import ascii_lowercase, ascii_uppercase
 

daal4py/dbscan.py

@@ -4,11 +4,9 @@
 
 import argparse
 from bench import (
-    parse_args, measure_function_time, load_data, print_output,
-    import_fptype_getter
+    parse_args, measure_function_time, load_data, print_output, getFPType
 )
 from daal4py import dbscan
-getFPType = import_fptype_getter()
 
 
 parser = argparse.ArgumentParser(description='daal4py DBSCAN clustering '

daal4py/df_clsf.py

@@ -5,14 +5,13 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, load_data, print_output, accuracy_score,
-    float_or_int, import_fptype_getter
+    float_or_int, getFPType
 )
 import numpy as np
 from daal4py import (
     decision_forest_classification_training,
     decision_forest_classification_prediction, engines_mt2203
 )
-getFPType = import_fptype_getter()
 
 
 def df_clsf_fit(X, y, n_classes, n_trees=100, seed=12345,

daal4py/df_regr.py

@@ -6,14 +6,13 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, load_data, print_output, rmse_score,
-    float_or_int, import_fptype_getter
+    float_or_int, getFPType
 )
 from daal4py import (
     decision_forest_regression_training,
     decision_forest_regression_prediction,
     engines_mt2203
 )
-getFPType = import_fptype_getter()
 
 
 def df_regr_fit(X, y, n_trees=100, seed=12345, n_features_per_node=0,

daal4py/distances.py

@@ -5,10 +5,9 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, print_output, load_data,
-    import_fptype_getter
+    getFPType
 )
 import daal4py
-getFPType = import_fptype_getter()
 
 
 def compute_distances(pairwise_distances, X):

daal4py/kmeans.py

@@ -4,12 +4,10 @@
 
 import argparse
 from bench import (
-    parse_args, measure_function_time, load_data, print_output,
-    import_fptype_getter
+    parse_args, measure_function_time, load_data, print_output, getFPType
 )
 import numpy as np
 from daal4py import kmeans
-getFPType = import_fptype_getter()
 
 
 parser = argparse.ArgumentParser(description='daal4py K-Means clustering '

daal4py/linear.py

@@ -5,10 +5,9 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, load_data, print_output, rmse_score,
-    import_fptype_getter
+    getFPType
 )
 from daal4py import linear_regression_training, linear_regression_prediction
-getFPType = import_fptype_getter()
 
 
 parser = argparse.ArgumentParser(description='daal4py linear regression '

daal4py/log_reg.py

@@ -5,14 +5,14 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, load_data, print_output, accuracy_score,
-    import_fptype_getter
+    getFPType
 )
 import numpy as np
 import daal4py
 from daal4py import math_logistic, math_softmax
 from daal4py.sklearn.utils import make2d
 import scipy.optimize
-getFPType = import_fptype_getter()
+
 
 _logistic_loss = daal4py.optimization_solver_logistic_loss
 _cross_entropy_loss = daal4py.optimization_solver_cross_entropy_loss

daal4py/pca.py

@@ -5,12 +5,12 @@
 import argparse
 from bench import (
     parse_args, measure_function_time, load_data, print_output,
-    import_fptype_getter
+    getFPType
 )
 import numpy as np
 from daal4py import pca, pca_transform, normalization_zscore
 from sklearn.utils.extmath import svd_flip
-getFPType = import_fptype_getter()
+
 
 parser = argparse.ArgumentParser(description='daal4py PCA benchmark')
 parser.add_argument('--svd-solver', type=str,