File size: 2,515 Bytes
6a18a2d ddeac02 6a18a2d ddeac02 01c59ec ddeac02 01c59ec ddeac02 44258d1 ddeac02 44258d1 f2f785a 44258d1 f2f785a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 |
---
dataset_info:
features:
- name: '#'
dtype: int64
- name: flower_diameter_cm
dtype: float64
- name: petal_length_cm
dtype: float64
- name: petal_width_cm
dtype: float64
- name: petal_count
dtype: int64
- name: stem_height_cm
dtype: float64
- name: color
dtype: string
splits:
- name: original
num_bytes: 1702
num_examples: 30
- name: augmented
num_bytes: 17020
num_examples: 300
download_size: 18466
dataset_size: 18722
configs:
- config_name: default
data_files:
- split: original
path: data/original-*
- split: augmented
path: data/augmented-*
pretty_name: f
---
# Dataset Card for Tabular Flower Data
### Dataset Details and Description
This dataset consists of 30 individual flowers. For each flower, five physical features were recorded: flower_diameter_cm (the width of the bloom at its widest point), petal_length_cm (average length of a petal), petal_width_cm (average width of a petal), petal_count (number of petals on the bloom), and stem_height_cm (approximate stem height from the soil line). The target variable is the color of the pedals, assigned categorically as Red, Yellow, Orange, Purple, Pink, or White. This small, structured dataset mimics the famous Iris dataset but with local flowers.
To increase the size of the dataset and provide variety for modeling, the original 30 rows were expanded to 300 synthetic samples using a stratified bootstrap with controlled noise. Specifically, rows were sampled with replacement within each color class to maintain the original class distribution. For each numeric feature, small random perturbations (Gaussian noise scaled to 10 % of the feature’s standard deviation) were added to simulate natural measurement variability. Values were then clamped to plausible ranges derived from the original data’s 1st–99th percentiles to avoid impossible numbers (like negative lengths), and integer-like features such as petal_count were rounded back to whole numbers. Categorical features were left unchanged so that each synthetic row retained a realistic label. This procedure produced a larger, more diverse dataset that still reflects realistic physical characteristics of the observed flowers.
### Dataset Intended Use
To practice AI ML
### Dataset Out of Scope
To predict flower properties.
### Personal and Sensitive Information
This dataset does not contain personal or sensitive information.
- **Curated by:** Scotty McGee
- **Language:** English
|