วีดีโอ

Hello! High correlation means that these two factors are very similar. Including them together doesn’t add value and may unnecessarily complicate the model.

Hello! Click the plus sign to the right of the "As guest" field. A credential input field will open, followed by a new window: 1. Enter any connection name you like. 2. Below, you can enter the Web GIS URL if desired. 3. Stay "Basic authentication" selected. 4. Then, enter your username and password.

UPD: Now you can update NextGIS Connect to version 2.6.0 and there will be a clearer authentication window.

Hello! Thank you for this question! 1. Code is opened and you can freely explore it: github.com/nextgis/qgis_molusce/tree/master/src/molusce 2. There is a built-in PDF documentation: github.com/nextgis/qgis_molusce/blob/master/src/molusce/doc/en/QuickHelp.pdf Take a look at chapter 3 with technical and algorithm details

Hello. In this particular case by two reasons: 1. Transforming DEM to metric coordinate reference system is crucial for calculating derivatives like slope angle 2. All input datasets should have the same extent, pixel size and coordinate reference system as input Land Use/Land Cover rasters, so we should transform everything to their spatial domain

Hard to say without data samples. You can send us some to support@nextgis.com and we'll try to figure out

Hello! It means that 59% of Barren pixels remained the same class (Barren). And 41% of them were transited to "Natural Herbaceous".

@@nextgis_channel Thanks very much for your answer 🙂

Hi, usually it means that your hypothesis about influecing environmental factors is wrong, or model parameters are wrong, or both. First step is to play a lot with model parameters, if metrics dont grow, then you should reconsider spatial variables

Hello! Write to the chat, please: t.me/nextgis_talks

There is a free plan, as well as paid ones: nextgis.com/pricing-base/

Yes

If "range min" and "range max" values are invalid, it seems to be a problem with your input rasters and their metadata, so plugin is unable to take correct min/max values from it

It's very simple and doesn't have many things to be tuned. You should only set up the process of converting continuous data to discrete, selecting number of intervals for each factor.

Hello! Actually hundreds of pixels were changed, so model was able to catch something. But in general hypothesis in this demo case was very weak, so I wouldn't count on results as something real. If we would take a look at transition potential maps for each pair of classes, we would notice that there many potential changes, but model wa not confident enough most of them, so it prefered to leave state unchanged.

Hello! In current implementation of MOLUSCE predictor maps are static for the whole period, including forecast. It is one of current limitations. Usually it's better to use predictors data relevant for the latest known LULC date. Resolution of data could be different, depending on your study scale. In this video we used MODIS MCD12Q1 data with ~500 meters/pixel spatial resolution.

@@akwijayanto Hi! First of all, the plugin was transited from the QGIS 2 engine to the QGIS 3 engine. It is a huge thing, a lot of internal mechanisms are changed. It also means that we are able to support Molusce again, which was nearly impossible before the update (too much outdated code was inside). Besides that: - Improved styles support. Now Molusce works with paletted/unique values render type, which is the actual standard for land cover rasters. In the previous version you had to use singleband pseudocolor, and in many cases (gaps in class numbers etc.) it wasn't convenient. - calculation efficiency of ANN method has been improved significantly, and stop button now works well (which is very important for fast experiments with training the model under different parameters) - some bugs and UI fixes

@@tahmidanam6391 hello, minimal QGIS version is 3.22

Hmm, set the quality to 1080, make the video full screen

Glad it worked for you!

Hello! To upload via the Connect module, you'll only need your data and/or a configured project in QGIS. Please ask any API-related questions in this chat: t.me/nextgis_talks

Satellite data, specifically cloud-free Sentinel 2 mosaics, can be obtained through the NextGIS Data service: data.nextgis.com/

Indeed, this tool generates a layer using the data from the table. It is important to note that each row should contain coordinates. In case you make any modifications to the table, in order for the new objects to appear on the map, you will need to utilize the tool again and choose the appropriate mode: "ADD" or "REPLACE".

it will depend on zoom scales needed

годится любая прямая публичная ссылка на любой опубликованный файл SVG

th-cam.com/video/JKMJ7DBZ-vs/w-d-xo.html

Нет проблем, вот тут есть примеры data.nextgis.com/ru/about/#pdf