Data Types

1
from kumoai import Dtype
2
3
# Numerical types
4
Dtype.bool      # Boolean values
5
Dtype.int       # Integer values
6
Dtype.float     # Floating point values
7
8
# String types
9
Dtype.string    # Text data
10
Dtype.binary    # Binary data
11
12
# Temporal types
13
Dtype.date      # Date/timestamp
14
Dtype.time      # Time values
15
Dtype.timedelta # Time differences
16
17
# List types
18
Dtype.floatlist   # Lists of floats (embeddings/sequences)
19
Dtype.intlist     # Lists of integers
20
Dtype.stringlist  # Lists of strings

1
import pandas as pd
2
import kumoai.rfm as rfm
3
4
df = pd.DataFrame({'user_id': [1, 2, 3]})
5
table = rfm.LocalTable(df, name="users")
6
7
print(table["user_id"].dtype)

1
from kumoai import Stype
2
3
# Core semantic types
4
Stype.numerical        # Numerical values for mathematical operations
5
Stype.categorical      # Discrete categories with limited cardinality
6
Stype.multicategorical # Multiple categories in single field
7
Stype.ID               # Unique identifiers
8
Stype.text             # Natural language text
9
Stype.timestamp        # Date/time information
10
Stype.sequence         # Embeddings or sequential data

1
# Check if a semantic type supports a data type
2
stype = Stype.categorical
3
dtype = Dtype.string
4
5
is_compatible = stype.supports_dtype(dtype)
6
print(f"{stype} supports {dtype}: {is_compatible}")
7
8
# Get default semantic type for a data type
9
default_stype = dtype.default_stype
10
print(f"Default stype for {dtype}: {default_stype}")

1
# Valid stype modifications (compatible with underlying dtype)
2
table['user_id'].stype = 'ID'           # Dtype.int -> Stype.ID
3
table['category'].stype = 'categorical' # Dtype.string -> Stype.categorical
4
table['description'].stype = 'text'     # Dtype.string -> Stype.text
5
6
# Invalid: integers do not support text semantic types:
7
table['user_id'].stype = 'text'