Issue: NDV Pixels in Change Map Generated by MOLUSCE Under area_changes > Create_change_map

[Prateekbhatt789]

Hello,

I am experiencing an issue while using the MOLUSCE plugin in QGIS, specifically when generating a change map using the Create_change_map function under the area_changes module.

The input rasters I am working with include:

2017_builtup
2020_builtup
DEM
Distance from road
Distance from waterbodies
When I attempt to generate the change map, the output raster contains several NDV (No Data Value) pixels, even though I have pixel values for all of these input rasters. These NDV pixels are affecting the analysis, and I am unable to correctly handle them in the generated change map.

Here are the steps I followed:

I used a vector layer representing the boundary of 2017 (AOI_boundary) to clip the DEM raster only.
I reprojected the DEM raster and the vector boundary to ensure they had matching CRS (EPSG:32644).
When running the Create_change_map function, the output raster contains NDV pixels, even though the input rasters and the factor rasters have valid data values in most areas.
I have attempted to use the Fill_no_data function, but the NDV issue persists in the change map.
I would appreciate any guidance on how to resolve this issue. It seems like the plugin might not be handling the NoData values properly when performing the change detection.

I look forward to your suggestions or any potential fixes. Thank you in advance!

Best regards,
Prateek.

[RemeshevskiyValeriy]

Hello, Prateek! Do you have the opportunity to share the dataset that you used when you encountered this issue in MOLUSCE? This would make it possible to find out the possible cause of the problem. You can try to send your dataset in this issue or to the email valeriy.remeshevskiy@nextgis.com

[Prateekbhatt789]

Subject- Dataset for Identifying Issue in MOLUSCE Dear Valeriy I hope this email finds you well. I am reaching out to seek your assistance in identifying the possible cause of an issue I encountered while using MOLUSCE. I have attached the dataset related to this issue for your reference. I would greatly appreciate it if you could review the dataset and provide valuable suggestions or insights to resolve the problem. Thank you for your time and support. I look forward to your response. Best regards, Prateek

…

On Mon, Dec 2, 2024 at 4:02 PM Valeriy ***@***.***> wrote: Hello, Prateek! Do you have the opportunity to share the dataset that you used when you encountered this issue in MOLUSCE? This would make it possible to find out the possible cause of the problem. You can try to send your dataset in this issue or to the email ***@***.*** — Reply to this email directly, view it on GitHub <#66 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASHEAHE7PP4EWNH5CNTLLAD2DQZKHAVCNFSM6AAAAABSU3UXHOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMJRGE2TGOBYGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[RemeshevskiyValeriy]

Prateek, hello again!

May I ask if you have sent your dataset to the email address specified in my previous message? If so, then, unfortunately, I did not receive your letter. If it doesn't bother you, then please try to send to this email - remeshevskiy@gmail.com . Hope that everything will work out :)

[Prateekbhatt789]

Hi valeriy, I apologize for replying late as i was busy with some different work. I am sending you my dataset again. Pls acknowledge it after receiving it.

…

On Thu, Dec 5, 2024 at 12:36 AM Valeriy ***@***.***> wrote: Prateek, hello again! May I ask if you have sent your dataset to the email address specified in my previous message? If so, then, unfortunately, I did not receive your letter. If it doesn't bother you, then please try to send to this email - ***@***.*** . Hope that everything will work out :) — Reply to this email directly, view it on GitHub <#66 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASHEAHAZG6AGRKC77XDXFHL2D5HEXAVCNFSM6AAAAABSU3UXHOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMJYGMZTEMJSGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[RemeshevskiyValeriy]

Hello, Prateek!

First of all, I apologize for keeping you waiting so long for updates about the problem you`re facing. The problem turned out to be not so innocuous.
MOLUSCE creates rasters using the GDAL library, and in cases where data blocks (in this case, raster strips) filled with zeros are contained in a row, the writing algorithm ignores these blocks by default thus interpreting them as "no data". A large number of users may have encountered this problem, so thank you for reporting this issue.
While the official version of the plugin with the bugfix has not been released, you can install MOLUSCE with problem fix from the zip archive which I attached to this message. I will be glad of any feedback.

Best wishes,
Valeriy

molusce-4.2.1.zip