CODEWITHSUNDEEP
pythonmatplotlibplot
Paul Stoner
Paul Stoner

Reputation: 1512

MATPLOTLIB - Increase scale for plot

I am working on an analysis of tornado data from 1996 to 2015. I am comparing the number of tornadoes to the total property loss by state.

I have a basic plot, but the smaller numbers are bunched too closely together

Tornado Plot

what I'd like to know is how I can scale this so as to spread these out a bit more?

My code for the plot is

count = tonadoes_1996_damage['count']
loss = tonadoes_1996_damage['total_loss']

max_tornado_count = tonadoes_1996_damage['count'].max()
max_tornado_cnt_label = tonadoes_1996_damage[tonadoes_1996_damage['count'] == max_tornado_count].index.tolist()[0]
max_tornado_cnt_x = tonadoes_1996_damage[tonadoes_1996_damage['count'] == max_tornado_count]['count']
max_tornado_cnt_y = tonadoes_1996_damage[tonadoes_1996_damage['count'] == max_tornado_count]['total_loss']

max_tornado_loss = tonadoes_1996_damage['total_loss'].max()
max_tornado_loss_label = tonadoes_1996_damage[tonadoes_1996_damage['total_loss'] == max_tornado_loss].index.tolist()[0]
max_tornado_loss_x = tonadoes_1996_damage[tonadoes_1996_damage['total_loss'] == max_tornado_loss]['count']
max_tornado_loss_y = tonadoes_1996_damage[tonadoes_1996_damage['total_loss'] == max_tornado_loss]['total_loss']

colors = np.random.rand(51)
area = count
plt.scatter(count, loss,s=area,c=colors,alpha=.5)

xlab = "Number of Tornadoes [in thousands]"
ylab = "Total Loss [in million USD]"
title = "Total Property Loss Since 1996"

plt.xlabel(xlab)
plt.xlim(0, 3500)

plt.ylabel(ylab)
plt.ylim(0, 6000)

plt.title(title)
plt.grid(True)

plt.text(max_tornado_cnt_x, max_tornado_cnt_y, max_tornado_cnt_label)
plt.text(max_tornado_loss_x, max_tornado_loss_y, max_tornado_loss_label)

plt.show()

I've played around with setting the x-ticks and y-ticks but those don't seem to adjust the scale, just add more ticks (unless I am doing something wrong.

Upvotes: 0

Views: 3586

Answers (1)

FLab
FLab

Reputation: 7466

You could use logarithmic scale for the x-axis and this will improve the visualisation: http://matplotlib.org/examples/pylab_examples/log_demo.html

Try something like:

fig = plt.figure(figsize=(11, 8))
ax = fig.add_subplot(1, 1, 1)
ax.set_xscale('log')
plt.scatter(count, loss,s=area,c=colors,alpha=.5)

and then add all properties accordingly.

Upvotes: 1

Related Questions