how to plot data in python from a file were the first column is a string and the second column is a number?

Question

Suppose I have a file called FARM with the following data

pigs 23
horses 10
cows 35
goats 20
sheep 22
chickens 62
roosters 5

I'd like to plot this data with the following result

The following code does exactly what I want

import matplotlib.pyplot as plt
import numpy as np
a,b = np.genfromtxt('FARM', usecols=(0,1), unpack=True, delimiter=' ', dtype='str')

x=[]
for i in a:
   x.append(i)

y=[]
for i in b:
   y.append(int(i))

plt.plot(x,y,'s')
plt.show()

My question is: is there a direct way to get x and y from np.genfromtxt, or do I have to convert from a and b as I have done?

If it were possible to set dtype selectively for each column, I'd have what I want. But I haven't been able to figure out how to do that.

After further thought, taking into account the comments and answers, I think a more intuitive way to plot columns with different datatypes from the same file is to make two calls to np.genfromtxt, each extracting a single column and then proceeding in the usual way, as follows:

x = np.genfromtxt('FARM', usecols=(0),  unpack=True,  dtype='str')                 
y = np.genfromtxt('FARM', usecols=(1),  unpack=True,  dtype='int')        

plt.plot(x,y,'s')
plt.show()

This way no special conversions are required.

Answer 1

There is a direct way to get the individual data types directly from the file. To do so, you have to use dtype=None. Doing so, the default dtypes will be retained. A list of tuples will be generated which you can split into two sublists x and y using zip and then directly plot.

data = np.genfromtxt('FARM', usecols=(0,1), delimiter=' ', dtype=None)
x, y = list(zip(*data))
plt.plot(x,y,'s')

An alternative to your approach would be directly to use map to convert strings to int instead of using a for loop.

a,b = np.genfromtxt('FARM', usecols=(0,1), unpack=True, delimiter=' ', dtype='str')

y = list(map(int, b))
plt.plot(a,y,'s')