# Fluid.mlpclassifier~: Wrong Point Size

**URL:** <https://discourse.flucoma.org/t/fluid-mlpclassifier-wrong-point-size/2251>\
**Category:** Usage Questions\
**Created:** [April 1, 2024, 9:27pm UTC](https://discourse.flucoma.org/t/fluid-mlpclassifier-wrong-point-size/2251 "2024-04-01T21:27:38Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![geosmina](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/geosmina/32/2324_2.png) [@geosmina](https://discourse.flucoma.org/u/geosmina)\
**Post date:** [April 1, 2024, 9:27pm UTC](https://discourse.flucoma.org/t/fluid-mlpclassifier-wrong-point-size/2251/1 "2024-04-01T21:27:38Z")

</div>

Hi. I’m using the `fluid.mlpclassifier~` object, and everything seems to work up until the usage of the ``predictpoint` message. I got 2 new buffers with test data, do I need to specify the exact sample size for the ML algorithm to work? I’ve been following this tutorial: [https://youtu.be/cjk9oHw7PQg?si=FKxXKPta9VFy3Ry8](https://youtu.be/cjk9oHw7PQg?si=FKxXKPta9VFy3Ry8)

```bash
fluid.mlpclassifier~: Wrong Point Size

```

This is how I have it, yo:

 ![image](https://discourse.flucoma.org/uploads/default/original/2X/a/a99697b514440e900d13f7eecd1883ac2ddcdcc3.png)

---

<div class="post-metadata">

**Author:** ![tremblap](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/tremblap/32/5_2.png) [@tremblap](https://discourse.flucoma.org/u/tremblap)\
**Post date:** [April 3, 2024, 12:46pm UTC](https://discourse.flucoma.org/t/fluid-mlpclassifier-wrong-point-size/2251/2 "2024-04-03T12:46:47Z")

</div>

Hello!

It looks ok but I wonder if there is not a point entered before textChico is sized. Can you print the dataset and labelset first? and send the full max window it might help us troubleshoot.