How to generate a CSV file for tiled?

[a416297338]

The CSV file located in the directory environment\frontend_server\static_dirs\assets\the_ville\matrix is used to assist in building the map. The "id" in this CSV file uses the "firstgid" from the .tmx file. I would like to know if the CSV file under "maze" is automatically generated, and whether the CSV file under "special_blocks" is handwritten after manual observation. If not, how were they created?

[wy12yw]

I also want to know the answer to this question, can you please solve it

[navindu-ds]

I also have some doubts to clarify on this matter. But let me give some ideas from what I understood.

As far as I understood, the environment image is rendered using the Tiled software, and the source file for the environment can be found at environment\frontend_server\static_dirs\assets\the_ville\visuals\the_ville.tmx. When attempting to export the .tmx file to the CSV format to be saved under the folder environment\frontend_server\static_dirs\assets\the_ville\matrix, there are 2 discrepancies identified between the CSV files exported by the Tiled software and the CSV files available in the frontend folder.

1. Difference in number of CSV files generated & used

The number of CSV files exported from Tiled was much higher compared to the five maze CSV files and five block CSV files in the repository.

The figure 1 shows the list of the fifteen CSV files exported from Tiled Software

The figure 2 shows the list of the five CSV files used by the Generative Agents repository

2. Difference in content in the CSV files

The content format of the CSV files differs as well. The provided CSV files in the repository predominantly contain 0's, while the files I generated contain -1's. It is assumed the predominant 0's or -1's represents the empty cells in the given layer.

The figure 3 shows the CSV exported from Tiled with mostly -1's

The figure 4 shows the CSV file available on the repo for the collision_matrix.csv filled with 0's instead

Conclusion

This discrepancy leads to believe that an additional process was applied to the CSV files during the export process, which is not indicated in the source code. It would be great if the authors (@joonspk-research) could provide some insight or documentation on any additional steps or transformations performed in this case.

[Kmaffitt]

I was thinking of making a new map and was left wondering the same thing.

from maze.py:
# Loading the maze. The mazes are taken directly from the json exports of # Tiled maps. They should be in csv format.

I exported the tiled project as json and grabbed the data array value from the Collisions object. After conversion to a single line and replacing all non 0 non 32125 values with 32125 it was identical to the included collision_matrix.csv. I have not yet compared the other included environment csv's and json export values but I imagine it's a similar story.

[pkmnjourney]

I was thinking of making a new map and was left wondering the same thing.

from maze.py: # Loading the maze. The mazes are taken directly from the json exports of # Tiled maps. They should be in csv format.

I exported the tiled project as json and grabbed the data array value from the Collisions object. After conversion to a single line and replacing all non 0 non 32125 values with 32125 it was identical to the included collision_matrix.csv. I have not yet compared the other included environment csv's and json export values but I imagine it's a similar story.

I have done the same and was able to get the characters behaving correctly within the new map. However, visually, it still shows the previous map. Were you able to get around that?