xtick label setting with improper scale

[fonnesbeck]

I have some data that I have used to generate a stacked bar graph.

by_conclusion = measles_data.groupby(["AGE", "CONCLUSION"])
counts_by_cause = by_conclusion.size().unstack().fillna(0)
ax = counts_by_cause.plot(kind='bar', stacked=True)

However, I get a pretty crowded x-axis:

So, I try to use set_xticks to try and thin them out, either as an argument to plot, or as a separate command:

ax.xticks(arange(0,80,10))

However, for some reason, the labels end up being downscaled by a factor of 10:

(might be hard to see, but it is printing "0.0, 1.0, 2.0, ..." instead of "0, 10, 20, ...")

Its not clear why this is happening, or how to get around it, so I'm either going about this wrong or its a bug. I get the same behavior when I try using MaxNLocator:

ax.xaxis.set_major_locator( MaxNLocator(nbins = 7) )

The data are here

[jorisvandenbossche]

Closing as duplicate of #7612

[jreback]

re-opening as not fixed by #26185

[NumberPiOso]

Data posted by @fonnesbeck is not available anymore

So I am trying to build a small reproducible example, I will leave it here in case it helps someone:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

n = 200
ages = np.random.choice(np.arange(0, 81), n)
outcome = np.random.choice(["Clinical", "Confirmed", "Discarded"], n)
df = pd.DataFrame({
    "conclusion": outcome,
    "age": ages
})
print(df)
#     conclusion  age
# 0    Discarded   74
# 1    Confirmed   63
# ...




counts = df.groupby(["conclusion", "age"]).size().unstack().fillna(0)
print(counts)
# age          0    1    2    3    4    5   ...   75   76   77   78   79   80
# conclusion                                ...                              
# Clinical    1.0  0.0  0.0  1.0  1.0  0.0  ...  1.0  2.0  0.0  2.0  1.0  0.0
# Confirmed   1.0  1.0  1.0  0.0  2.0  1.0  ...  0.0  1.0  1.0  0.0  1.0  1.0
# Discarded   0.0  1.0  1.0  1.0  1.0  2.0  ...  1.0  0.0  1.0  0.0  0.0  1.0


ax = counts.plot(kind='bar', stacked=True)
plt.show()

However the result is quite different in pandas version 1.5.0.dev0+305.g6a5cdf585d

[NumberPiOso]

So I am trying to build a completely different example, more minimalistic to show the error, but everything is working fine for me:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
n = 80
df = pd.DataFrame({"Clinical": np.random.choice([0, 1, 2, 3], n),
                   "Confirmed": np.random.choice([0, 1, 2, 3], n),
                   "Discarded": np.random.choice([0, 1, 2, 3], n),
                   }, index=np.arange(0, n))

ax = df.plot(kind='bar', stacked=True)

produces

if we try:

ax.xticks(np.arange(0,80,10))
AttributeError: 'AxesSubplot' object has no attribute 'xticks'

but using

ax.set_xticks(np.arange(0,80,10))

will produce the desired result

So I am not sure if this issue should be marked as information needed or if it is already solutioned. @jreback and @mroeschke.

[mroeschke]

@NumberPiOso thanks for investigating. If all appears well for your minimal example might be good to add your example as a unit test.