Why do I get "ufunc 'multiply' did not contain a loop with signature matching types dtype('S32') dtype('S32') dtype('S32')" with values from raw_input

Question

I am trying to create a really simple program that will plot a plot a parabola where v is velocity, a is acceleration and x is time. The user will input values for v and a, then v and a and x will determine y.

I attempted to do this with this:

x = np.linspace(0., 9., 10)
a = raw_input('Acceleration =')
v = raw_input('Velocity = ')
y = v * x - 0.5 * a * x**2.

But, I keep getting this error:

TypeError: ufunc 'multiply' did not contain a loop with signature matching types dtype('S32') dtype('S32') dtype('S32')

What does this mean?

Solution 1

From the documentation of raw_input:

The function then reads a line from input, converts it to a string (stripping a trailing newline), and returns that.

So what happens is that you try to multiply a string with a float, something like y="3" * x - 0.5 * "3" *x**2, which is not defined.

The easiest way to circumvent this is to cast the input string to float first.

x = np.linspace(0., 9., 10)
a = float(raw_input('Acceleration ='))
v = float(raw_input('Velocity = '))
y = v * x - 0.5 * a * x**2

Mind that if you're using Python 3, you'd need to use input instead of raw_input,

a = float(input('Acceleration ='))

Solution 2

I faced this problem recently, change the dtype of x to something specific by doing:

x = np.asarray(x, dtype='float64')

Comments

[AsheKetchum]

why is that the easiest way?

[ImportanceOfBeingErnest]

Well, for python2 you can use input instead of raw_input but that hides a bit the origin of the problem. And since input evaluates the input, there are all kind of hidden traps, while float(raw_input()) either works or throws an error.

[AsheKetchum]

so casting to float is the safest way?

[ImportanceOfBeingErnest]

Safe in the sense that it may fail and throw an error, but you will always know why it fails (because you haven't typed in a number). For beginners it's definitely the most understandable way.