refactor(data preprocess): remove the cut off options from info.json (#…

…200)

* refactor(data preprocess): remove the cut off options from info.json and collect the values from input.json

* update LMDB info.json. not need anymore.

* refactor(default_dataset): refactor the _TrajData for ase data.

Previous the ase data will be transferred into text file and then loaded by the _TrajData. now i refactor the function.

both text and ase data are treated equally. will works as a class funtion to initial the _TrajData class.

* add print logo in main and format some of  the logger.info

* update argcheck  collect_cutoffs.  add new function with  get_cutoffs_from_model_options .

* Fix(get_cutoffs_from_model_options) : fix rcut in  powerlaw and varTang96.

For powerlaw and varTang96, the rs is not exactly the hard cutoff. so when extract the r_max for data. we have to use rs + 5 * w; but for other method just use rs.

* update band post process.

* update test

* update test

* update build and get_cutoffs_from_model_options to support the rmax to be dict.

* refactor(build dataset): change build_dataset from function to a class instance and add from_model class function.

note, compared to the previous build_dataset, this one is more flexible.
previous build_dataset is a function. now i define a class DataBuilder and re-defined __call__ function.  then build_dataset is an instance of DataBuilder class. so i can use build_dataset.from_model() to build dataset from model. at the same time the previous way to use  build_dataset is still available. like build_dataset(...).

* add checkcutoff in dataset builder.

* update AtomicData_options to make it compatible with older versions

* Update argcheck.py

---------

Co-authored-by: Yinzhanghao Zhou <64253517+floatingCatty@users.noreply.github.com>

.gitignore

@@ -6,6 +6,9 @@ dptb/tests/**/*.pth
 dptb/tests/**/*.npy
 dptb/tests/**/*.traj
 dptb/tests/**/out*/*
+dptb/tests/**/out*/*
+dptb/tests/**/*lmdb
+dptb/tests/**/*h5
 examples/_*
 *.dat
 *log*

dptb/__main__.py

@@ -1,5 +1,39 @@
-from dptb.entrypoints.main import main
+from dptb.entrypoints.main import main as entry_main
+import logging
+import pyfiglet
+from dptb import __version__
 
+logging.basicConfig(level=logging.INFO, format='%(message)s')
+log = logging.getLogger(__name__)
+
+def print_logo():
+    f = pyfiglet.Figlet(font='dos_rebel')  # 您可以选择您喜欢的字体
+    logo = f.renderText("DeePTB")
+    log.info(" ")
+    log.info(" ")
+    log.info("#"*81)
+    log.info("#" + " "*79 + "#")
+    log.info("#" + " "*79 + "#")
+    for line in logo.split('\n'):
+        if line.strip():  # 避免记录空行
+            log.info('#     '+line+ '     #')
+    log.info("#" + " "*79 + "#")
+    version_info = f"Version: {__version__}"
+    padding = (79 - len(version_info)) // 2
+    nspace = 79-padding
+    format_str = "#" + "{}"+"{:<"+f"{nspace}" + "}"+ "#"         
+    log.info(format_str.format(" "*padding, version_info))
+    log.info("#" + " "*79 + "#")
+    log.info("#"*81)
+    log.info(" ")
+    log.info(" ")
+def main() -> None:
+    """
+    The main entry point for the dptb package.
+    """
+    print_logo()
+    entry_main()
 
 if __name__ == '__main__':
-    main()
+    #print_logo()
+    main()

dptb/data/AtomicData.py

@@ -496,7 +496,7 @@ def from_points(
     def from_ase(
         cls,
         atoms,
-        r_max,
+        r_max: Union[float, int, dict],
         er_max: Optional[float] = None,
         oer_max: Optional[float] = None,
         key_mapping: Optional[Dict[str, str]] = {},