# Percent slicing and rounding + Stratify

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fpercent-slicing-and-rounding-stratify%2F43588%3E%5C
**Category:** Beginners\
**Created:** [June 17, 2023, 7:04am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fpercent-slicing-and-rounding-stratify%2F43588 "2023-06-17T07:04:43Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![ltkw98](https://avatars.discourse-cdn.com/v4/letter/l/74df32/32.png) [@ltkw98](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fltkw98)%5C
**Post date:** [June 17, 2023, 7:04am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fpercent-slicing-and-rounding-stratify%2F43588%2F1 "2023-06-17T07:04:43Z")

</div>

I want to incorporate 4 fold cross validation, however how do I ensure the train and validation are stratified?

Codes:  
from datasets import load\_dataset

# The first 75% of dataset

train\_75\_25pct\_ds = load\_dataset(‘dataset’, split=‘train[:75%]’)  
train\_75\_25pct\_ds

# The last 25% of dataset

validation\_75\_25pct\_ds = load\_dataset(‘dataset’, split=‘train[-25%:]’)  
validation\_75\_25pct\_ds

---

<div class="post-metadata">

**Author:** ![mariosasko](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/mariosasko/32/31548_2.png) [@mariosasko](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fmariosasko)%5C
**Post date:** [June 19, 2023, 3:52pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fpercent-slicing-and-rounding-stratify%2F43588%2F2 "2023-06-19T15:52:29Z")

</div>

You can do this as follows:

```python
train_ds = load_dataset(‘dataset’, split=‘train’)
ddict = train_ds.train_test_split(train_size=0.25, stratify_by_column="col_name")
train_ds, validation_ds = ddict["train"], ddict["test"]

```