Comments (15)
Just to follow up on the above, those reference were very useful in helping tweak parameters. When I did modify the parameters, both steps seemed to complete successfully
from igneous.
Hi vikramcns,
Thanks for writing in! The floating point bounds on the skeleton file is probably the culprit. Can you share with me your info file? That would help me better understand what is going on.
Will
from igneous.
yes! the contents of the info file in the skeletons_mip_0
folder are below:
{"@type": "neuroglancer_skeletons", "transform": [1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0], "vertex_attributes": [{"id": "radius", "data_type": "float32", "num_components": 1}], "spatial_index": {"resolu
tion": [4200.0, 4200.0, 4200.0], "chunk_size": [2150400.0, 2150400.0, 2150400.0]}, "mip": 0}
from igneous.
Interesting, can you also show me the info file for the main image volume?
from igneous.
Yes, see below. The ellipses indicate downsampled scales:
{
"data_type": "uint64",
"mesh": "mesh_mip_0_err_40",
"num_channels": 1,
"scales": [
{
"chunk_sizes": [
[
128,
128,
128
]
],
"compressed_segmentation_block_size": [
8,
8,
8
],
"encoding": "compressed_segmentation",
"key": "4200.0_4200.0_4200.0",
"resolution": [
4200.0,
4200.0,
4200.0
],
"size": [
4904,
4904,
2946
],
"voxel_offset": [
0,
0,
0
]
},...
],
"skeletons": "skeletons_mip_0",
"type": "segmentation"
}
from igneous.
Okay, so I think that on cursory inspection, handling these floating point names is a feature I should modify cloudvolume to support. Give me a bit, I think I almost have a fix.
from igneous.
Perfect, thank you!
from igneous.
Okay, try installing pip install cloud-volume==8.9.2
and see if that fixes things. There could be other problems that crop up, but that one should not be the issue anymore.
from igneous.
I will test that now and get back to you shortly
from igneous.
seems to be working so far! I will let you know if anything breaks
from igneous.
so after a few hours it looks like there is no progress on the actual tasks. the progress bar shows 0/999 after 3 hours with parallel=16
for the fusing skeletons step
from igneous.
You may have selected parameters during initial skeletonization that are not appropriate for your dataset. If there are loops or many tiny branches introduced, this can result in very slow merging (with poor final results too). You can also try killing the parallel job and seeing if a single core makes any progress. If its taking up too much memory and your working with swap, that could lead to very slow progress.
Give this tutorial a look, it might help you debug your run:
- https://github.com/seung-lab/kimimaro/wiki/A-Pictorial-Guide-to-TEASAR-Skeletonization
- https://github.com/seung-lab/kimimaro/wiki/Intuition-for-Setting-Parameters-const-and-scale
from igneous.
from igneous.
from igneous.
Closing due to inactivity. Reopen if you need it!
from igneous.
Related Issues (20)
- Variable name bug HOT 1
- cloudfiles.exceptions.DecompressionError: File contains zero bytes: N/A HOT 4
- Examples of using mulitres mesh? HOT 6
- Mesh info file caching causing errors in MultiResShardedMeshMergeTask? HOT 2
- Rearrange Slices Operator
- feat: rotate queues
- uncompressed sharded format HOT 3
- Last voxel column gets shifted to the front HOT 6
- How do I export the resulting skeleton to a swc file? HOT 5
- Some volumes are not Downsampled HOT 2
- Does igneous currently support generating graphene format data HOT 1
- Why is the browser stuck when visualizing the generated surface mesh? HOT 7
- `num_mips` needs a type annotation HOT 3
- Check datasets are identical.
- Do I need to manually clear the folder where QUEUE is stored? HOT 2
- How to avoid the cloudvolume.exceptions.EmptyVolumeException HOT 1
- Troubleshooting with ptq HOT 11
- Image create command crashes for path including "." HOT 2
- Data generated with `igneous image create` loads slowly in NG HOT 11
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from igneous.