rootloader 1.3.0__tar.gz → 1.4.0__tar.gz

rootloader-1.4.0/.claude/settings.local.json

@@ -0,0 +1,14 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(python -m py_compile rootloader/th2.py rootloader/tleaf.py rootloader/ttree.py rootloader/tdirectory.py)",
+      "Bash(python3 -m py_compile rootloader/th2.py rootloader/tleaf.py rootloader/ttree.py rootloader/tdirectory.py)",
+      "Bash(python -m pytest tests/ -x --tb=short -q)",
+      "Bash(python3 -m pytest tests/ --tb=short -q)",
+      "Bash(python3 -m pytest tests/ --tb=no -q)",
+      "Bash(python3 -m pytest tests/test_ttree.py::test_stats_scalar_single_column_float tests/test_ttree.py::test_loc_slice_reduces_size --tb=long -q)",
+      "Bash(python3 -m pytest tests/ -v --tb=short -q)",
+      "Bash(python3 -m pytest tests/ -v --tb=short)"
+    ]
+  }
+}

rootloader-1.4.0/CLAUDE.md

@@ -0,0 +1,31 @@
+# CLAUDE.md
+
+## Project Overview
+
+`rootloader` is a Python library that wraps PyROOT to load ROOT files into memory as Python-native objects (numpy arrays, pandas DataFrames). It targets physicists at TRIUMF who work with ROOT data but prefer Python workflows.
+
+## Architecture
+
+The class hierarchy mirrors ROOT's object model:
+
+```
+attrdict (dict subclass — attribute-style key access)
+└── tdirectory   — wraps ROOT.TDirectoryFile or ROOT.TFile; iterates keys and dispatches to typed wrappers
+    └── tfile    — entry point; opens a ROOT.TFile and delegates to tdirectory
+
+ttree            — wraps ROOT.TTree via RDataFrame; supports lazy column selection, filters, and conversion to pandas/numpy
+th1              — wraps ROOT.TH1; stores x/y/dy arrays and provides .plot() and .to_dataframe()
+th2              — wraps ROOT.TH2; stores x/y/z/dz arrays and provides .plot() and .to_dataframe()
+tleaf            — thin wrapper around ROOT.TLeaf for direct leaf-level access
+```
+
+**Key design decisions:**
+
+- `attrdict` makes all dict keys accessible as attributes (`fid.tree1` == `fid['tree1']`). This is the base for `tdirectory`.
+- `tdirectory.__init__` resolves ROOT key cycles (keeps highest cycle number) and dispatches each object to `ttree`, `th1`, `th2`, or a nested `tdirectory`.
+- `ttree` is lazy: it holds a `ROOT.RDataFrame` and only materializes data when `.to_dataframe()`, `.to_dict()`, `.to_array()`, or a stats method (`.min()`, `.max()`, etc.) is called. Column selection via `__getitem__` returns a view (not a copy).
+- Stats methods (`min`, `max`, `mean`, `sum`, `std`) use JIT-compiled C++ templates (`cpp_template` in `ttree.py`) to avoid memory creep from repeated ROOT JIT compilations. Each unique `(function, dtype)` pair is compiled once and cached on the ROOT module.
+- `th1.to_dataframe()` and `th2.to_dataframe()` store reconstruction metadata in `df.attrs` so the original typed object can be restored via `tdirectory.from_dataframe()`.
+- `ROOT.EnableImplicitMT()` is called at import time in `ttree.py` to enable multithreading for RDataFrame operations.
+- `ROOT.gROOT.SetBatch(1)` is set at import in `__init__.py` to suppress graphical output.
+

{rootloader-1.3.0 → rootloader-1.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rootloader
-Version: 1.3.0
+Version: 1.4.0
 Summary: Read simple root files into memory
 Project-URL: Homepage, https://github.com/ucn-triumf/rootloader
 Project-URL: Bug Tracker, https://github.com/ucn-triumf/rootloader/issues
@@ -687,6 +687,9 @@ Requires-Dist: matplotlib
 Requires-Dist: numpy
 Requires-Dist: pandas
 Requires-Dist: tqdm
+Provides-Extra: test
+Requires-Dist: pytest; extra == 'test'
+Requires-Dist: pytest-cov; extra == 'test'
 Description-Content-Type: text/markdown
 
 # ROOTLOADER

{rootloader-1.3.0 → rootloader-1.4.0}/docs/rootloader/tdirectory.md

@@ -41,7 +41,7 @@ class tdirectory(attrdict):
 
 ### tdirectory.copy
 
-[Show source in tdirectory.py:122](../../rootloader/tdirectory.py#L122)
+[Show source in tdirectory.py:127](../../rootloader/tdirectory.py#L127)
 
 Make a copy of this object
 
@@ -53,7 +53,7 @@ def copy(self): ...
 
 ### tdirectory.from_dataframe
 
-[Show source in tdirectory.py:131](../../rootloader/tdirectory.py#L131)
+[Show source in tdirectory.py:136](../../rootloader/tdirectory.py#L136)
 
 Convert all elements contained in self to original objects
 
@@ -65,7 +65,7 @@ def from_dataframe(self): ...
 
 ### tdirectory.to_dataframe
 
-[Show source in tdirectory.py:142](../../rootloader/tdirectory.py#L142)
+[Show source in tdirectory.py:147](../../rootloader/tdirectory.py#L147)
 
 Convert all objects possible (th1, th2, and ttree) into pandas dataframes
 

{rootloader-1.3.0 → rootloader-1.4.0}/docs/rootloader/th2.md

@@ -75,7 +75,7 @@ def plot(self, ax=None, flat=True, **kwargs): ...
 
 ### th2.to_dataframe
 
-[Show source in th2.py:159](../../rootloader/th2.py#L159)
+[Show source in th2.py:160](../../rootloader/th2.py#L160)
 
 Convert tree to pandas dataframe
 

{rootloader-1.3.0 → rootloader-1.4.0}/docs/rootloader/ttree.md

@@ -4,6 +4,12 @@
 
 > Auto-generated documentation for [rootloader.ttree](../../rootloader/ttree.py) module.
 
+#### Attributes
+
+- `cpp_template` - template to pre-compile stats functions to avoid memory creep during JIT compilations
+  used in ttree._get_stat: '\n{TYPE2} RDF_{FNNAME}_{TYPE2}({DTYPE} df, string col){\n    return df.{FNNAME}<{TYPE2}>(col).GetValue.\n}\n'
+
+
 - [ttree](#ttree)
   - [ttree](#ttree-1)
     - [ttree.__getitem__](#ttree__getitem__)
@@ -19,17 +25,19 @@
     - [ttree.min](#ttreemin)
     - [ttree.reset](#ttreereset)
     - [ttree.reset_columns](#ttreereset_columns)
-    - [ttree.rms](#ttreerms)
     - [ttree.set_filter](#ttreeset_filter)
     - [ttree.set_index](#ttreeset_index)
     - [ttree.size](#ttreesize)
     - [ttree.std](#ttreestd)
+    - [ttree.sum](#ttreesum)
+    - [ttree.to_array](#ttreeto_array)
     - [ttree.to_dataframe](#ttreeto_dataframe)
     - [ttree.to_dict](#ttreeto_dict)
+    - [ttree.values](#ttreevalues)
 
 ## ttree
 
-[Show source in ttree.py:14](../../rootloader/ttree.py#L14)
+[Show source in ttree.py:22](../../rootloader/ttree.py#L22)
 
 Extract ROOT.TTree with lazy operation. Looks like a dataframe in most ways
 
@@ -43,12 +51,12 @@ Extract ROOT.TTree with lazy operation. Looks like a dataframe in most ways
 
 ```python
 class ttree(object):
-    def __init__(self, tree): ...
+    def __init__(self, tree, filter_string=None, columns=None): ...
 ```
 
 ### ttree.__getitem__
 
-[Show source in ttree.py:79](../../rootloader/ttree.py#L79)
+[Show source in ttree.py:90](../../rootloader/ttree.py#L90)
 
 Fetch a new dataframe with fewer 'columns', as a memory view
 
@@ -60,7 +68,9 @@ def __getitem__(self, key): ...
 
 ### ttree.columns
 
-[Show source in ttree.py:318](../../rootloader/ttree.py#L318)
+[Show source in ttree.py:345](../../rootloader/ttree.py#L345)
+
+Return list of column (branch) names
 
 #### Signature
 
@@ -71,7 +81,9 @@ def columns(self): ...
 
 ### ttree.filters
 
-[Show source in ttree.py:321](../../rootloader/ttree.py#L321)
+[Show source in ttree.py:349](../../rootloader/ttree.py#L349)
+
+Return list of RDataFrame filters
 
 #### Signature
 
@@ -82,7 +94,7 @@ def filters(self): ...
 
 ### ttree.hist1d
 
-[Show source in ttree.py:129](../../rootloader/ttree.py#L129)
+[Show source in ttree.py:139](../../rootloader/ttree.py#L139)
 
 Return histogram of column
 
@@ -107,7 +119,7 @@ def hist1d(self, column=None, nbins=None, step=None, edges=None): ...
 
 ### ttree.hist2d
 
-[Show source in ttree.py:182](../../rootloader/ttree.py#L182)
+[Show source in ttree.py:192](../../rootloader/ttree.py#L192)
 
 Return histogram of two columns
 
@@ -131,7 +143,9 @@ def hist2d(
 
 ### ttree.index
 
-[Show source in ttree.py:324](../../rootloader/ttree.py#L324)
+[Show source in ttree.py:353](../../rootloader/ttree.py#L353)
+
+Return ttree of just the index data
 
 #### Signature
 
@@ -142,7 +156,9 @@ def index(self): ...
 
 ### ttree.index_name
 
-[Show source in ttree.py:327](../../rootloader/ttree.py#L327)
+[Show source in ttree.py:357](../../rootloader/ttree.py#L357)
+
+Return string of the name of the index branch
 
 #### Signature
 
@@ -153,7 +169,9 @@ def index_name(self): ...
 
 ### ttree.loc
 
-[Show source in ttree.py:330](../../rootloader/ttree.py#L330)
+[Show source in ttree.py:361](../../rootloader/ttree.py#L361)
+
+Return a ttree that can be indexed like a pandas dataframe
 
 #### Signature
 
@@ -164,7 +182,9 @@ def loc(self): ...
 
 ### ttree.max
 
-[Show source in ttree.py:354](../../rootloader/ttree.py#L354)
+[Show source in ttree.py:402](../../rootloader/ttree.py#L402)
+
+Return the max value of the tree, for each branch
 
 #### Signature
 
@@ -174,7 +194,9 @@ def max(self): ...
 
 ### ttree.mean
 
-[Show source in ttree.py:359](../../rootloader/ttree.py#L359)
+[Show source in ttree.py:405](../../rootloader/ttree.py#L405)
+
+Return the mean value of the tree, for each branch
 
 #### Signature
 
@@ -184,7 +206,9 @@ def mean(self): ...
 
 ### ttree.min
 
-[Show source in ttree.py:349](../../rootloader/ttree.py#L349)
+[Show source in ttree.py:399](../../rootloader/ttree.py#L399)
+
+Return the min value of the tree, for each branch
 
 #### Signature
 
@@ -194,7 +218,7 @@ def min(self): ...
 
 ### ttree.reset
 
-[Show source in ttree.py:254](../../rootloader/ttree.py#L254)
+[Show source in ttree.py:264](../../rootloader/ttree.py#L264)
 
 Make a new tree
 
@@ -206,7 +230,7 @@ def reset(self): ...
 
 ### ttree.reset_columns
 
-[Show source in ttree.py:258](../../rootloader/ttree.py#L258)
+[Show source in ttree.py:268](../../rootloader/ttree.py#L268)
 
 Include all columns again
 
@@ -216,19 +240,9 @@ Include all columns again
 def reset_columns(self): ...
 ```
 
-### ttree.rms
-
-[Show source in ttree.py:364](../../rootloader/ttree.py#L364)
-
-#### Signature
-
-```python
-def rms(self): ...
-```
-
 ### ttree.set_filter
 
-[Show source in ttree.py:268](../../rootloader/ttree.py#L268)
+[Show source in ttree.py:278](../../rootloader/ttree.py#L278)
 
 Set a filter on the dataframe to select a subset of the data
 
@@ -240,7 +254,7 @@ def set_filter(self, expression, inplace=False): ...
 
 ### ttree.set_index
 
-[Show source in ttree.py:262](../../rootloader/ttree.py#L262)
+[Show source in ttree.py:272](../../rootloader/ttree.py#L272)
 
 Set the index column name
 
@@ -252,7 +266,9 @@ def set_index(self, column): ...
 
 ### ttree.size
 
-[Show source in ttree.py:333](../../rootloader/ttree.py#L333)
+[Show source in ttree.py:365](../../rootloader/ttree.py#L365)
+
+Return the number of rows in the ttree
 
 #### Signature
 
@@ -263,7 +279,9 @@ def size(self): ...
 
 ### ttree.std
 
-[Show source in ttree.py:369](../../rootloader/ttree.py#L369)
+[Show source in ttree.py:411](../../rootloader/ttree.py#L411)
+
+Return the standard deviationif the of values the tree, for each branch
 
 #### Signature
 
@@ -271,11 +289,35 @@ def size(self): ...
 def std(self): ...
 ```
 
+### ttree.sum
+
+[Show source in ttree.py:408](../../rootloader/ttree.py#L408)
+
+Return the sum of the values of the tree, for each branch
+
+#### Signature
+
+```python
+def sum(self): ...
+```
+
+### ttree.to_array
+
+[Show source in ttree.py:289](../../rootloader/ttree.py#L289)
+
+Return ttree data as 1D or 2D numpy array (depending on number of columns)
+
+#### Signature
+
+```python
+def to_array(self): ...
+```
+
 ### ttree.to_dataframe
 
-[Show source in ttree.py:280](../../rootloader/ttree.py#L280)
+[Show source in ttree.py:300](../../rootloader/ttree.py#L300)
 
-Return pandas dataframe of the data
+Return ttree data as pandas dataframe
 
 #### Signature
 
@@ -285,10 +327,25 @@ def to_dataframe(self): ...
 
 ### ttree.to_dict
 
-[Show source in ttree.py:307](../../rootloader/ttree.py#L307)
+[Show source in ttree.py:334](../../rootloader/ttree.py#L334)
+
+Return ttree data as dict of numpy arrays
 
 #### Signature
 
 ```python
 def to_dict(self): ...
+```
+
+### ttree.values
+
+[Show source in ttree.py:369](../../rootloader/ttree.py#L369)
+
+Convert ttree 1D or 2D numpy array (depending on number of columns)
+
+#### Signature
+
+```python
+@property
+def values(self): ...
 ```

{rootloader-1.3.0 → rootloader-1.4.0}/pyproject.toml

@@ -22,6 +22,9 @@ dynamic = ["version"]
 "Homepage" = "https://github.com/ucn-triumf/rootloader"
 "Bug Tracker" = "https://github.com/ucn-triumf/rootloader/issues"
 
+[project.optional-dependencies]
+test = ["pytest", "pytest-cov"]
+
 # set version
 [tool.hatch.version]
 path = "rootloader/version.py"

{rootloader-1.3.0 → rootloader-1.4.0}/rootloader/tdirectory.py

@@ -65,7 +65,12 @@ class tdirectory(attrdict):
             # TTree
             if 'TTree' == classname:
                 if empty_ok or obj.GetEntries() > 0:
-                    self[name] = ttree(obj)
+                    if name in tree_filter:
+                        filter_string, columns = tree_filter[name]
+                        self[name] = ttree(obj, filter_string=filter_string,
+                                           columns=columns)
+                    else:
+                        self[name] = ttree(obj)
                 elif not quiet:
                     tqdm.write(f'Skipped "{name}" due to lack of entries')
 

{rootloader-1.3.0 → rootloader-1.4.0}/rootloader/th2.py

@@ -71,7 +71,7 @@ class th2(object):
         self.dz = self.dz.transpose()
 
     def __len__(self):
-        return self.nbins
+        return self.nbinsx * self.nbinsy
 
     def __repr__(self):
         return f'{self.base_class}: "{self.name}", {self.entries} entries, sum = {self.sum}'
@@ -143,7 +143,8 @@ class th2(object):
 
             if ax is None:
                 plt.figure()
-                ax = plt.gca().add_subplot(projection='3d')
+                # add_subplot is a Figure method, not an Axes method
+                ax = plt.gcf().add_subplot(111, projection='3d')
 
             ax.plot_surface(xx, yy, self.z, **kwargs)
 

{rootloader-1.3.0 → rootloader-1.4.0}/rootloader/tleaf.py

@@ -16,7 +16,7 @@ class tleaf(object):
 
     def copy(self):
         """Make a copy of this object"""
-        return tleaf(self._leaf.copy())
+        return tleaf(self._leaf)
 
     def get_entry(self, i):
         self._leaf.GetBranch().GetEntry(i)