Skip to content

Commit cf507c6

Browse files
MartinPdeSMartinPdeS
committed
adding example and fixing Workflow class
1 parent 0215cc3 commit cf507c6

3 files changed

Lines changed: 199 additions & 21 deletions

File tree

‎FlowCyPy/workflow.py‎

Lines changed: 34 additions & 20 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,36 @@
11
from typing import List
2+
from TypedUnit import (
3+
Length,
4+
Power,
5+
FlowRate,
6+
Dimensionless,
7+
Frequency,
8+
Resistance,
9+
Time,
10+
)
11+
from FlowCyPy import SimulationSettings
212

313
from pydantic import ConfigDict
414
from pydantic.dataclasses import dataclass
515

6-
from FlowCyPy import distribution, units
7-
from FlowCyPy.circuits import SignalProcessor
8-
from FlowCyPy.flow_cytometer import FlowCytometer
16+
from FlowCyPy.units import ureg
17+
from FlowCyPy.fluidics import distributions # noqa: F401
18+
from FlowCyPy.fluidics import population # noqa: F401
19+
from FlowCyPy.sampling_method import GammaModel, ExplicitModel # noqa: F401
920
from FlowCyPy.fluidics import FlowCell, Fluidics, ScattererCollection
21+
22+
23+
from FlowCyPy.flow_cytometer import FlowCytometer
1024
from FlowCyPy.opto_electronics import (
1125
Detector,
1226
GaussianBeam,
1327
OptoElectronics,
1428
TransimpedanceAmplifier,
1529
)
16-
from FlowCyPy.peak_locator.base_class import BasePeakLocator
30+
from FlowCyPy.binary.peak_locator import BasePeakLocator
31+
1732
from FlowCyPy.population import Sphere
33+
from FlowCyPy import classifier as classifiers
1834
from FlowCyPy.signal_processing import (
1935
Digitizer,
2036
SignalProcessing,
@@ -30,31 +46,30 @@
3046
@dataclass(config=config_dict, kw_only=True)
3147
class Workflow:
3248
# Source parameters
33-
wavelength: units.Quantity
34-
source_numerical_aperture: units.Quantity
35-
optical_power: units.Quantity
49+
wavelength: Length
50+
source_numerical_aperture: Dimensionless
51+
optical_power: Power
3652

3753
# Flowcell parameters
38-
sample_volume_flow: units.Quantity
39-
sheath_volume_flow: units.Quantity
40-
width: units.Quantity
41-
height: units.Quantity
54+
sample_volume_flow: FlowRate
55+
sheath_volume_flow: FlowRate
56+
width: Length
57+
height: Length
4258

4359
# Opto-electronic parameters
4460
detectors: List[Detector] = None
4561
bit_depth: str
4662
saturation_levels: str
47-
sampling_rate: units.Quantity
48-
background_power: units.Quantity = 0 * units.watt
63+
sampling_rate: Frequency
64+
background_power: Power = 0 * ureg.watt
4965

5066
# Population parameters
5167
populations: List[Sphere]
52-
medium_refractive_index: units.Quantity
5368

5469
# signal processing parameters
55-
gain: units.Quantity
56-
bandwidth: units.Quantity
57-
analog_processing: List[SignalProcessor] = None
70+
gain: Resistance
71+
bandwidth: Frequency
72+
analog_processing: List[object] = None
5873
peak_locator: BasePeakLocator
5974
trigger: BaseTrigger
6075

@@ -73,7 +88,6 @@ def _get_fluidics(self) -> Fluidics:
7388
Fluidics
7489
"""
7590
scatterer_collection = ScattererCollection(
76-
medium_refractive_index=self.medium_refractive_index,
7791
populations=self.populations,
7892
)
7993

@@ -145,5 +159,5 @@ def initialize(self) -> None:
145159
background_power=self.background_power,
146160
)
147161

148-
def run(self, run_time: units.Quantity):
149-
self.results = self.cytometer.run(run_time=run_time)
162+
def run(self, run_time: Time):
163+
return self.cytometer.run(run_time=run_time)

‎docs/examples/tutorials/workflow_1.py‎

Lines changed: 164 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,164 @@
1+
"""
2+
Flow Cytometry Simulation: Full System Example with Workflow
3+
============================================================
4+
5+
This tutorial demonstrates a complete flow cytometry simulation using the FlowCyPy library.
6+
It models fluidics, optics, signal processing, and classification of multiple particle populations.
7+
8+
Steps Covered:
9+
--------------
10+
1. Configure simulation parameters and noise models
11+
2. Define laser source, flow cell geometry, and fluidics
12+
3. Add synthetic particle populations
13+
4. Set up detectors, amplifier, and digitizer
14+
5. Simulate analog and digital signal acquisition
15+
6. Apply triggering and peak detection
16+
7. Classify particle events based on peak features
17+
"""
18+
19+
from FlowCyPy.workflow import (
20+
ureg,
21+
SimulationSettings,
22+
Workflow,
23+
Detector,
24+
circuits,
25+
peak_locator,
26+
triggering_system,
27+
distributions,
28+
population,
29+
GammaModel,
30+
classifiers,
31+
)
32+
33+
SimulationSettings.include_noises = False
34+
SimulationSettings.include_shot_noise = True
35+
SimulationSettings.include_dark_current_noise = True
36+
SimulationSettings.include_source_noise = True
37+
SimulationSettings.include_amplifier_noise = True
38+
SimulationSettings.assume_perfect_hydrodynamic_focusing = True
39+
SimulationSettings.population_cutoff_bypass = False
40+
41+
population_0 = population.Sphere(
42+
name="Pop 0",
43+
medium_refractive_index=distributions.Delta(1.33 * ureg.RIU),
44+
concentration=5e10 * ureg.particle / ureg.milliliter,
45+
diameter=distributions.RosinRammler(
46+
shape=150 * ureg.nanometer,
47+
scale=50 * ureg.nanometer,
48+
low_cutoff=50.0 * ureg.nanometer,
49+
),
50+
refractive_index=distributions.Normal(
51+
mean=1.44 * ureg.RIU,
52+
standard_deviation=0.002 * ureg.RIU,
53+
low_cutoff=1.33 * ureg.RIU,
54+
),
55+
)
56+
57+
population_1 = population.Sphere(
58+
name="Pop 1",
59+
medium_refractive_index=distributions.Delta(1.33 * ureg.RIU),
60+
concentration=5e17 * ureg.particle / ureg.milliliter,
61+
diameter=distributions.RosinRammler(
62+
shape=50 * ureg.nanometer,
63+
scale=50 * ureg.nanometer,
64+
),
65+
refractive_index=distributions.Normal(
66+
mean=1.44 * ureg.RIU,
67+
standard_deviation=0.002 * ureg.RIU,
68+
low_cutoff=1.33 * ureg.RIU,
69+
),
70+
sampling_method=GammaModel(mc_samples=10_000),
71+
)
72+
73+
74+
detector_0 = Detector(
75+
name="side",
76+
phi_angle=90 * ureg.degree,
77+
numerical_aperture=0.3 * ureg.AU,
78+
responsivity=1 * ureg.ampere / ureg.watt,
79+
)
80+
81+
detector_1 = Detector(
82+
name="forward",
83+
phi_angle=0 * ureg.degree,
84+
numerical_aperture=0.3 * ureg.AU,
85+
responsivity=1 * ureg.ampere / ureg.watt,
86+
)
87+
88+
discriminator = triggering_system.DynamicWindow(
89+
trigger_detector_name="forward",
90+
threshold="2sigma",
91+
pre_buffer=20,
92+
post_buffer=20,
93+
)
94+
95+
peak_locator = peak_locator.GlobalPeakLocator(compute_width=False)
96+
97+
analog_processing = [
98+
circuits.BaselineRestorator(window_size=10 * ureg.microsecond),
99+
circuits.BesselLowPass(cutoff=2 * ureg.megahertz, order=4, gain=2),
100+
]
101+
102+
workflow = Workflow(
103+
wavelength=405 * ureg.nanometer,
104+
source_numerical_aperture=0.1 * ureg.AU,
105+
optical_power=200 * ureg.milliwatt,
106+
sample_volume_flow=80 * ureg.microliter / ureg.minute,
107+
sheath_volume_flow=1 * ureg.milliliter / ureg.minute,
108+
width=200 * ureg.micrometer,
109+
height=100 * ureg.micrometer,
110+
populations=[population_0, population_1],
111+
gain=10 * ureg.volt / ureg.ampere,
112+
bandwidth=10 * ureg.megahertz,
113+
bit_depth="14bit",
114+
sampling_rate=60 * ureg.megahertz,
115+
saturation_levels="auto",
116+
background_power=0.001 * ureg.milliwatt,
117+
detectors=[detector_0, detector_1],
118+
analog_processing=analog_processing,
119+
trigger=discriminator,
120+
peak_locator=peak_locator,
121+
)
122+
123+
workflow.initialize()
124+
125+
run_record = workflow.run(run_time=1 * ureg.millisecond)
126+
127+
_ = run_record.event_collection.plot(x="Diameter")
128+
129+
# %%
130+
# Step 5: Plot Events and Raw Analog Signals
131+
# ------------------------------------------
132+
_ = run_record.event_collection.plot(x="forward")
133+
134+
135+
# %%
136+
# Plot raw analog signals
137+
# -----------------------
138+
_ = run_record.plot_analog(figure_size=(12, 8))
139+
140+
141+
# %%
142+
# Step 6: Plot Triggered Analog Segments
143+
# --------------------------------------
144+
_ = run_record.plot_digital(figure_size=(12, 8))
145+
146+
147+
# %%
148+
# Step 7: Plot Peak Features
149+
# --------------------------
150+
_ = run_record.peaks.plot(x=("forward", "Height"))
151+
152+
153+
# %%
154+
# Step 8: Classify Events from Peak Features
155+
# ------------------------------------------
156+
classifier = classifiers.KmeansClassifier(number_of_clusters=2)
157+
158+
classified = classifier.run(
159+
dataframe=run_record.peaks.unstack("Detector"),
160+
features=["Height"],
161+
detectors=["side", "forward"],
162+
)
163+
164+
_ = classified.plot(x=("side", "Height"), y=("forward", "Height"))

‎pyproject.toml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -73,7 +73,7 @@ keywords = [
7373
]
7474

7575

76-
76+
7777
[project.optional-dependencies]
7878
testing = [
7979
"pytest>=7",

0 commit comments

Comments
 (0)