# Prior predicitive for linear regression
**URL:** https://discourse.pymc.io/t/prior-predicitive-for-linear-regression/3304
**Category:** Questions
**Created:** [May 21, 2019, 6:50am UTC](https://discourse.pymc.io/t/prior-predicitive-for-linear-regression/3304 "2019-05-21T06:50:54Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@Rahul\_Deora](https://discourse.pymc.io/u/Rahul_Deora)
#### Post date: [May 21, 2019, 6:50am UTC](https://discourse.pymc.io/t/prior-predicitive-for-linear-regression/3304/1 "2019-05-21T06:50:54Z")
I am unable to get the prior predictive distribution. Here is the reproducible code:
```
df = pd.read_csv('https://raw.githubusercontent.com/aloctavodia/Statistical-Rethinking-with-Python-and-PyMC3/master/Data/Howell1.csv',sep=';')
with pm.Model() as m4_3:
alpha = pm.Normal('alpha', mu=178, sd=100)
beta = pm.Normal('beta', mu=0, sd=10)
sigma = pm.Uniform('sigma', lower=0, upper=50)
mu = pm.Deterministic('mu', alpha + beta * (df.weight - df.weight.mean() )) # try uncomenting this line and comenting the above line
height = pm.Normal('height', mu=mu, sd=sigma)
trace_4_3 = pm.sample(1000, tune=1000)
```
I get
```
The error when converting the test value to that variable type:
Wrong number of dimensions: expected 0, got 1 with shape (544,).
```
beause I have put in a input of weight in beta, but not output of height. How do I generate Prior predicitive for this model?
---
### Author:  [@Rahul\_Deora](https://discourse.pymc.io/u/Rahul_Deora)
#### Post date: [May 22, 2019, 7:51am UTC](https://discourse.pymc.io/t/prior-predicitive-for-linear-regression/3304/2 "2019-05-22T07:51:59Z")
Can anyone help me with this?
---
### Author:  [@chartl](https://discourse.pymc.io/u/chartl)
#### Post date: [May 22, 2019, 3:40pm UTC](https://discourse.pymc.io/t/prior-predicitive-for-linear-regression/3304/3 "2019-05-22T15:40:22Z")
```auto
with pm.Model() as m4_3:
alpha = pm.Normal('alpha', mu=178, sd=100)
beta = pm.Normal('beta', mu=0, sd=10)
sigma = pm.Uniform('sigma', lower=0, upper=50)
mu = pm.Deterministic('mu', alpha + beta * (df.weight - df.weight.mean() )) # try uncomenting this line and comenting the above line
height = pm.Normal('height', mu=mu, sd=sigma, testval=100.)
trace_4_3 = pm.sample_prior_predictive(1000)
vn = [x for x in trace_4_3.keys() if x[-1] != '_']
for v in vn:
if len(trace_4_3[v].shape) > 1:
sbn.kdeplot(trace_4_3[v][:,0])
else:
sbn.kdeplot(trace_4_3[v])
```
