lets-plot 4.6.0rc1__cp38-cp38-macosx_10_15_x86_64.whl

lets_plot/plot/series_meta.py

@@ -0,0 +1,129 @@
+#  Copyright (c) 2024. JetBrains s.r.o.
+#  Use of this source code is governed by the MIT license that can be found in the LICENSE file.
+from datetime import datetime
+from typing import Union, Dict, Iterable
+
+from lets_plot._type_utils import is_polars_dataframe
+from lets_plot.plot.util import is_pandas_data_frame
+
+TYPE_INTEGER = 'int'
+TYPE_FLOATING = 'float'
+TYPE_STRING = 'str'
+TYPE_BOOLEAN = 'bool'
+TYPE_DATE_TIME = 'datetime'
+TYPE_UNKNOWN = 'unknown'
+
+
+def infer_type(data: Union[Dict, 'pandas.DataFrame', 'polars.DataFrame']) -> Dict[str, str]:
+    type_info = {}
+
+    if is_pandas_data_frame(data):
+        import pandas as pd
+        import numpy as np  # np is a dependency of pandas, we can import it without checking
+
+        for var_name, var_content in data.items():
+            if data.empty:
+                type_info[var_name] = TYPE_UNKNOWN
+                continue
+
+            inferred_type = pd.api.types.infer_dtype(var_content.values, skipna=True)
+            if inferred_type == "categorical":
+                dtype = var_content.cat.categories.dtype
+
+                if np.issubdtype(dtype, np.integer):
+                    type_info[var_name] = TYPE_INTEGER
+                elif np.issubdtype(dtype, np.floating):
+                    type_info[var_name] = TYPE_FLOATING
+                elif np.issubdtype(dtype, np.object_):
+                    # Check if all elements are strings
+                    if all(isinstance(x, str) for x in var_content.cat.categories):
+                        type_info[var_name] = TYPE_STRING
+                    else:
+                        type_info[var_name] = TYPE_UNKNOWN
+                else:
+                    type_info[var_name] = TYPE_UNKNOWN
+            else:
+                # see https://pandas.pydata.org/docs/reference/api/pandas.api.types.infer_dtype.html
+                if inferred_type == 'string':
+                    type_info[var_name] = TYPE_STRING
+                elif inferred_type == 'floating':
+                    type_info[var_name] = TYPE_FLOATING
+                elif inferred_type == 'integer':
+                    type_info[var_name] = TYPE_INTEGER
+                elif inferred_type == 'boolean':
+                    type_info[var_name] = TYPE_BOOLEAN
+                elif inferred_type == 'datetime64' or inferred_type == 'datetime':
+                    type_info[var_name] = TYPE_DATE_TIME
+                elif inferred_type == "date":
+                    type_info[var_name] = TYPE_DATE_TIME
+                elif inferred_type == 'empty':  # for columns with all None values
+                    type_info[var_name] = TYPE_UNKNOWN
+                else:
+                    type_info[var_name] = 'unknown(pandas:' + inferred_type + ')'
+    elif is_polars_dataframe(data):
+        import polars as pl
+        from polars.datatypes.group import INTEGER_DTYPES, FLOAT_DTYPES
+        for var_name, var_type in data.schema.items():
+
+            # https://docs.pola.rs/api/python/stable/reference/datatypes.html
+            if var_type in FLOAT_DTYPES:
+                type_info[var_name] = TYPE_FLOATING
+            elif var_type in INTEGER_DTYPES:
+                type_info[var_name] = TYPE_INTEGER
+            elif var_type == pl.datatypes.String:
+                type_info[var_name] = TYPE_STRING
+            elif var_type == pl.datatypes.Boolean:
+                type_info[var_name] = TYPE_BOOLEAN
+            elif var_type == pl.datatypes.Date or var_type == pl.datatypes.Datetime:
+                type_info[var_name] = TYPE_DATE_TIME
+            else:
+                type_info[var_name] = 'unknown(polars:' + str(var_type) + ')'
+    elif isinstance(data, dict):
+        for var_name, var_content in data.items():
+            if isinstance(var_content, Iterable):
+                if not any(True for _ in var_content):  # empty
+                    type_info[var_name] = TYPE_UNKNOWN
+                    continue
+
+                type_set = set(type(val) for val in var_content)
+                if type(None) in type_set:
+                    type_set.remove(type(None))
+
+                if len(type_set) == 0:
+                    continue
+
+                if len(type_set) > 1:
+                    if all(issubclass(type_obj, int) or issubclass(type_obj, float) for type_obj in type_set):
+                        type_info[var_name] = TYPE_FLOATING
+                    else:
+                        type_info[var_name] = 'unknown(mixed types)'
+                    continue
+
+                try:
+                    import numpy
+                except ImportError:
+                    numpy = None
+
+                type_obj = list(type_set)[0]
+                if type_obj == bool:
+                    type_info[var_name] = TYPE_BOOLEAN
+                elif issubclass(type_obj, int):
+                    type_info[var_name] = TYPE_INTEGER
+                elif issubclass(type_obj, float):
+                    type_info[var_name] = TYPE_FLOATING
+                elif issubclass(type_obj, str):
+                    type_info[var_name] = TYPE_STRING
+                elif issubclass(type_obj, datetime):
+                    type_info[var_name] = TYPE_DATE_TIME
+                elif numpy and issubclass(type_obj, numpy.datetime64):
+                    type_info[var_name] = TYPE_DATE_TIME
+                elif numpy and issubclass(type_obj, numpy.timedelta64):
+                    type_info[var_name] = TYPE_DATE_TIME
+                elif numpy and issubclass(type_obj, numpy.integer):
+                    type_info[var_name] = TYPE_INTEGER
+                elif numpy and issubclass(type_obj, numpy.floating):
+                    type_info[var_name] = TYPE_FLOATING
+                else:
+                    type_info[var_name] = 'unknown(python:' + str(type_obj) + ')'
+
+    return type_info