Custom CRS give the wrong result on Proj4Dart.

[pbrobo]

Now I am using Proj4Dart V2.0.0 & Flutter 3.0.4, found some big different of transform coordinates from my custom CRS. This is my code:

`import 'package:proj4dart/proj4dart.dart';

void main() {
var n1 = 1657193.5292;
var e1 = 834934.4417;
var ptSrc = Point(x: e1, y: n1);
var projSrc =
Projection.parse('+proj=utm +zone=47 +datum=WGS84 +units=m +no_defs');
var projDst = Projection.parse(
'+proj=tmerc +lat_0=0.0 +lon_0=102.25 +k_0=1.0 +x_0=500000 +y_0=0 +a=6378297.0 +b=6356911.77779 +towgs84=0.0,0.0,0.0,0,0,0,0 +units=m +no_defs type=crs');
var tuple = ProjectionTuple(fromProj: projSrc, toProj: projDst);
var ptDst = tuple.forward(ptSrc);
print(
'FORWARD: Transform point ${ptSrc.toArray()} from EPSG:32647 to custom projection: ${ptDst.toArray()}');
}
`
The console printed:
FORWARD: Transform point [834934.4417, 1657193.5292] from EPSG:32647 to custom projection: [485333.96880976285, 1655549.4642448912]

The correct answer must be [485333.9690, 1655549.7316].
This is very strange due to I used pyproj version of Python that derived from PROJ4 also but gave the correct coordinates.
This is my Python code with the same Proj4 strings:
`from pyproj import CRS, Transformer

projsrc = CRS.from_proj4('+proj=utm +zone=47 +datum=WGS84 +units=m +no_defs')
projdst = CRS.from_proj4('+proj=tmerc +lat_0=0.0 +lon_0=102.25 +k_0=1.0 +x_0=500000 +y_0=0 +a=6378297.0 +b=6356911.77779 +towgs84=0.0,0.0,0.0,0,0,0,0 +units=m +no_defs type=crs');
xx, yy = 834934.4417, 1657193.5292
transformer = Transformer.from_crs(projsrc, projdst)
ptdst = transformer.transform(xx=xx, yy=yy)
print(f'Source point x,y: {xx} {yy}')
print(f'Dest point x,y: {ptdst[0]} {ptdst[1]}')`

The console of Python printed:
Source point x,y: 834934.4417 1657193.5292
Dest point x,y: 485333.9689740725 1655549.7315796772

I am new to Dart/Flutter what I am doing wrong or is it Proj4Dart's bug? Please help me. I run this code on Windows 11.

[maRci002]

I'm sorry to inform you but this is the correct result since this project is based on proj4js which results the same.

var script = document.createElement("script");
var scriptPath = "https://cdnjs.cloudflare.com/ajax/libs/proj4js/2.8.0/proj4.js";
script.setAttribute("src", scriptPath);
document.head.appendChild(script);

script.onload = function() {
    proj4.defs('projSrc', '+proj=utm +zone=47 +datum=WGS84 +units=m +no_defs');
    proj4.defs('projDst', '+proj=tmerc +lat_0=0.0 +lon_0=102.25 +k_0=1.0 +x_0=500000 +y_0=0 +a=6378297.0 +b=6356911.77779 +towgs84=0.0,0.0,0.0,0,0,0,0 +units=m +no_defs type=crs');
    var result = proj4('projSrc', 'projDst').forward([834934.4417, 1657193.5292]);
    console.log(result);
}

output: [485333.96880976285, 1655549.4642448917]

You can run the above script in your browser's console (note: you cannot run on every site since this is script injection and some sites like github prevents it, so go for example to http://proj4js.org/ and press F12 and paste / run the above code snippet)

Anyway Python might use slightly different projection algorithm or Python might use higher/lower precision while calculating, the results almost the same I think the difference is still lower than 5mm which is insignificant if you think about it, it's a much bigger problem that the continents move from year to year.