flightanalysis 0.3.3__tar.gz → 0.3.4__tar.gz

{flightanalysis-0.3.3/flightanalysis.egg-info → flightanalysis-0.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: flightanalysis
-Version: 0.3.3
+Version: 0.3.4
 Summary: A package for analysing flight data
 Author-email: Thomas David <thomasdavid0@gmail.com>
 License: GNU GPL v3

{flightanalysis-0.3.3 → flightanalysis-0.3.4}/flightanalysis/scoring/criteria/f3a_criteria.py

@@ -20,7 +20,7 @@ class F3AIntra:
     track = Continuous(Exponential.fit_points(np.radians([30, 90]), [1.75, 6], 6))
     roll = Continuous(Exponential.fit_points(np.radians([30, 90]), [1.25, 6], 6))
     radius = Continuous(Exponential.fit_points([1.5, 3], [0.5, 1], 1))
-    speed = ContinuousValue(Exponential.fit_points([5, 15], [0.03, 0.09], 0.5))
+    speed = ContinuousValue(Exponential.fit_points([5, 15], [0.1, 0.3], 0.5))
     roll_rate = Continuous(Exponential.fit_points([1, 3], [0.02, 0.06], 0.5))
     autorotation_rate = Continuous(Exponential.fit_points([1, 3], [0.02, 0.06], 0.5))
     stallturn_speed = Limit(Exponential.fit_points([2, 4], [0.05, 0.1], 4), 4)

{flightanalysis-0.3.3 → flightanalysis-0.3.4}/flightanalysis/scoring/downgrade.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 from flightdata import Collection, State
-from .criteria import Bounded, Continuous, Single, Criteria
+from .criteria import Bounded, Continuous, Single, Criteria, ContinuousValue
 from .measurement import Measurement
 from .visibility import visibility
 from .results import Results, Result
@@ -11,7 +11,6 @@ from flightanalysis.base.ref_funcs import RefFuncs, RefFunc
 import numpy as np
 
 
-
 @dataclass
 class DownGrade:
     """This is for Intra scoring, it sits within an El and defines how errors should be measured and the criteria to apply
@@ -27,7 +26,9 @@ class DownGrade:
     ]  # gets the flown and template measurements
     smoothers: RefFuncs  # smoothes the measurement
     selectors: RefFuncs  # selects the values to downgrade
-    criteria: (Bounded | Continuous | Single)  # looks up the downgrades based on the errors
+    criteria: (
+        Bounded | Continuous | Single
+    )  # looks up the downgrades based on the errors
     display_name: str
 
     def rename(self, name: str):
@@ -52,9 +53,14 @@ class DownGrade:
 
     def __call__(self, fl, tp, limits=True) -> Result:
         measurement: Measurement = self.measure(fl, tp)
-        
-        sample = visibility(self.criteria.prepare(measurement.value), measurement.visibility, self.criteria.lookup.error_limit)
-        
+
+        sample = visibility(
+            self.criteria.prepare(measurement.value),
+            measurement.visibility,
+            self.criteria.lookup.error_limit,
+            'deviation' if isinstance(self.criteria, ContinuousValue) else 'value',
+        )
+
         for sm in self.smoothers:
             sample = sm(sample)
 
@@ -65,14 +71,14 @@ class DownGrade:
             fl = State(fl.data.iloc[ids])
             sample = sample[sub_ids]
             ids = ids[sub_ids]
-        
+
         return Result(
             self.display_name,
             measurement,
             sample,
             ids,
             *self.criteria(sample, limits),
-            self.criteria
+            self.criteria,
         )
 
 
@@ -84,7 +90,9 @@ def dg(
     selectors: RefFunc | list[RefFunc],
     criteria: Criteria,
 ):
-    return DownGrade(name, measure, RefFuncs(smoothers), RefFuncs(selectors) , criteria, display_name)
+    return DownGrade(
+        name, measure, RefFuncs(smoothers), RefFuncs(selectors), criteria, display_name
+    )
 
 
 class DownGrades(Collection):

flightanalysis-0.3.4/flightanalysis/scoring/visibility.py

@@ -0,0 +1,31 @@
+import numpy as np
+import pandas as pd
+
+def visibility(val, factor: float, limit: float, kind: str = 'value'):
+    """factor between 0 and 1"""
+
+    b = 1.8 - factor * 0.8
+    if kind == 'value':
+        norm = np.abs(val / limit)
+        return np.where(norm > 1, norm, norm**b) * limit * np.sign(val)
+    elif kind=='deviation':
+        diff = np.insert(np.diff(val), 0, 0.0, axis=0)
+        norm = np.abs(diff / limit)
+
+        res = np.where(norm > 1, norm, norm**b) * limit * np.sign(diff) 
+        return res.cumsum() + val[0]
+    else:
+        raise ValueError(f'kind {kind} not recognized')
+
+
+
+
+if __name__=='__main__':
+    import plotly.express as px
+
+
+    x = np.linspace(0, 20, 100)
+    px.line(pd.DataFrame({k: visibility(x, k, 10) for k in np.linspace(0.1,1,9)}, index=x)).show()
+
+    x = np.linspace(0, 2, 100)
+    px.line(pd.DataFrame({k: visibility(x, k, 1) for k in np.linspace(0.1,1,9)}, index=x)).show()

{flightanalysis-0.3.3 → flightanalysis-0.3.4/flightanalysis.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: flightanalysis
-Version: 0.3.3
+Version: 0.3.4
 Summary: A package for analysing flight data
 Author-email: Thomas David <thomasdavid0@gmail.com>
 License: GNU GPL v3

flightanalysis-0.3.3/flightanalysis/scoring/visibility.py

@@ -1,21 +0,0 @@
-import numpy as np
-import pandas as pd
-
-def visibility(val, factor: float, limit: float):
-    """factor between 0 and 1"""
-    norm = np.abs(val / limit)
-    b = 2 - factor
-    return np.where(norm > 1, norm, norm**b) * limit * np.sign(val)
-
-
-
-
-if __name__=='__main__':
-    import plotly.express as px
-
-
-    x = np.linspace(0, 20, 100)
-    px.line(pd.DataFrame({k: visibility(x, k, 10) for k in np.linspace(0.1,1,9)}, index=x)).show()
-
-    x = np.linspace(0, 2, 100)
-    px.line(pd.DataFrame({k: visibility(x, k, 1) for k in np.linspace(0.1,1,9)}, index=x)).show()