# Estimating T1 decoherence times

Fitters are used to estimate the T1 time based on experiment results from `t1_circuits`

executed on noisy devices. The estimate is based on the probability formula of measuring `1`

from the following equation, where *A*, *T**1*, and *B* are unknown parameters:

Since we set the T1 value earlier when we defined the noise model of the qubit as `20`

, let's assume for now that we do not know that and initialize the value to some percentage value away from the actual value. The `T1Fitter`

class has a few parameters that it needs in order to characterize the qubit. We will start by initializing the values for `a`

, `t1`

, and `b`

:

```
# Initialize the parameters for the T1Fitter, A, T1, and B
param_t1 = t1*1.2
param_a = 1.0
param_b = 0.0
```

This will initialize our `t1`

, `a`

, and `b`

parameters, which we will use to generate `T1Fitter`

.

Next, we will generate `T1Fitter`

by providing the following parameters:

`backend_result`

: The results from our test circuits on the backend...