# Suggestions for netcdf data format

**URL:** <https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462>\
**Category:** Data\
**Created:** [May 24, 2023, 5:14pm UTC](https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462 "2023-05-24T17:14:44Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![gradoj](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/gradoj/32/2273_2.png) [@gradoj](https://discourse.pangeo.io/u/gradoj)\
**Post date:** [May 24, 2023, 5:14pm UTC](https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462/1 "2023-05-24T17:14:44Z")

</div>

i am having a little trouble getting my data in the right format and would love any suggestions. I have radio stations scattered about which broadcast periodically are detected and recorded by multiple other radio sites. I started with H.2.5 Indexed Ragged Array of Timeseries because the station example is very close but it isn’t quite right for what i’m doing.

Each rf broadcast contains a unique id that identifies that transmission so i am thinking more about using something like the trajectory examples cause i’d like to be able to see which station sent the data and which stations received the transmission. I’m just not sure exactly how to structure it. I know the lat,lng for where it was sent and lat,lng from each station it was received at with metadata. Those could look like many trajectories? allowing me to fetch data based on the txid potentially? Any help appreciated.

the data looks something like this:

> {  
> txid: ‘123456’  
> txReport: {  
> time: 1244353545  
> stationid: 12324234  
> lat: 30.0  
> lon: 40.0  
> }  
> rxReports: {  
> 0: {  
> time: 1244353546  
> stationid: 64545343  
> lat: 31.0  
> lon: 41.0  
> data1: 17  
> }  
> 1: {  
> time: 1244353544  
> stationid: 23435454  
> lat: 32.0  
> lon: 43.0  
> data1: 16  
> }  
> …  
> }  
> }

---

<div class="post-metadata">

**Author:** ![ThomasMGeo](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/thomasmgeo/32/1722_2.png) [@ThomasMGeo](https://discourse.pangeo.io/u/ThomasMGeo)\
**Post date:** [May 26, 2023, 1:49pm UTC](https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462/2 "2023-05-26T13:49:44Z")

</div>

Is the data gridded? Why not just a table/csv/parquet file?

---

<div class="post-metadata">

**Author:** ![gradoj](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/gradoj/32/2273_2.png) [@gradoj](https://discourse.pangeo.io/u/gradoj)\
**Post date:** [May 26, 2023, 4:38pm UTC](https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462/3 "2023-05-26T16:38:23Z")

</div>

No, it is not gridded.  
There are a couple of TB of binary data so i’m really trying to get to zarr or similar so it can be a little more usable.

---

<div class="post-metadata">

**Author:** ![gradoj](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/gradoj/32/2273_2.png) [@gradoj](https://discourse.pangeo.io/u/gradoj)\
**Post date:** [May 28, 2023, 12:35am UTC](https://discourse.pangeo.io/t/suggestions-for-netcdf-data-format/3462/4 "2023-05-28T00:35:28Z")

</div>

Thanks for the recommendation. It looks like parquet will do what I need with a larger than memory dataset