# matlab – Initialize empty matrix in Python

## matlab – Initialize empty matrix in Python

If you are using `numpy`

arrays, you initialize to 0, by specifying the expected matrix size:

```
import numpy as np
d = np.zeros((2,3))
>>> d
[[ 0. 0. 0.]
[ 0. 0. 0.]]
```

This would be the equivalent of MATLAB s:

```
d = zeros(2,3);
```

You can also initialize an empty array, again using the expected dimensions/size

```
d = np.empty((2,3))
```

If you are not using numpy, the closest somewhat equivalent to MATLABs `d = []`

(i.e., a zero-size matrix) would be using an empty list and then

append values (for filling a vector)

```
d = []
d.append(0)
d.append(1)
>>> d
[0, 1]
```

or append lists (for filling a matrix row or column):

```
d = []
d.append(range(0,2))
d.append(range(2,4))
>>> d
[[0, 1], [2, 3]]
```

See also:

NumPy array initialization (fill with identical values) (SO)

How do I create an empty array/matrix in NumPy? (SO)

You could use a nested list comprehension:

```
# size of matrix n x m
matrix = [ [ 0 for i in range(n) ] for j in range(m) ]
```

#### matlab – Initialize empty matrix in Python

What about initializing a list, populating it, then converting to an array.

```
demod4 = []
```

Or, you could just populate at initialization using a list comprehension

```
demod4 = [[func(i, j) for j in range(M)] for i in range(N)]
```

Or, you could initialize an array of all zeros if you know the size of the array ahead of time.

```
demod4 = [[0 for j in range(M)] for i in range(N)]
```

or

```
demod4 = [[0 for i in range(M)]*N]
```

Or try using `numpy`

.

```
import numpy as np
N, M = 100, 5000
np.zeros((N, M))
```