# boulderopalscaleup.experiments.ReadoutClassifier

Source: https://docs.q-ctrl.com/references/boulder-opal-scale-up/boulderopalscaleup/experiments/ReadoutClassifier

```python
class boulderopalscaleup.experiments.ReadoutClassifier(*, transmon, recycle_delay_ns=200000, shot_count=5000, measure_waveform=None, classifier_state_count=2, use_active_reset=False, run_mixer_calibration=False, update='auto')
```

Parameters for training a readout classifier.

This does not optimize the readout pulse itself.
The measure waveform is fixed (provided or using the device’s default).
To optimize the readout pulse use ReadoutOptimization.

### Variables
* **transmon** (*str*) – The reference for the transmon to target.
* **recycle_delay_ns** (*int* *,* *optional*) – The delay between consecutive shots, in nanoseconds. Defaults to 200,000 ns.
* **shot_count** (*int* *,* *optional*) – The number of shots to take. Defaults to 5,000.
* **measure_waveform** ([*ConstantWaveform*](https://docs.q-ctrl.com/references/boulder-opal-scale-up/boulderopalscaleup/common/ConstantWaveform.md#boulderopalscaleup.common.ConstantWaveform) *or* *None* *,* *optional*) – The waveform to use for the measurement pulse.
* **classifier_state_count** (*int* *,* *optional*) – 

    The number of states to classify. If greater than 2, assumes that the higher
    : energy levels were previously calibrated. Currently only supports 2 or 3 states.
      Defaults to 2.

    Defaults to the measurement defcal.
  * **use_active_reset** (*bool* *,* *optional*) – Whether to enable active reset for 2-state classifier training when the stored
    active reset fidelity permits it. Defaults to False, using passive reset with
    recycle_delay_ns, so that training does not depend on the existing classifier.
    3-state training always uses passive reset.
  * **run_mixer_calibration** (*bool* *,* *optional*) – Whether to run mixer calibrations before running a program. Defaults to False.
  * **update** ( *"auto"* *or*  *"off"* *or*  *"prompt"* *,* *optional*) – How the device should be updated after an experiment run. Defaults to auto.
