问题
I wanted to do some plots based on imported csv files from a chemical analysis. So I import as follows:
In [91]:
df = pd.read_csv('/file_location/Untitled 1.csv', delimiter = '\;', index_col = 'IR')
df
Out[91]:
Sample 1 Sample 2
IR
300 1 0
400 5 4
500 6 0
600 0 8
4 rows × 2 columns
In [98]:
df.plot()
Fine looks good.
By convention this type of data i plotted with the x axis in descending order. Highest number to the right (do not ask me why). So i reorder the index-col:
In [97]:
df2 = df.sort_index(axis=0, ascending=False, kind='quicksort')
df2
Out[97]:
Sample 1 Sample 2
IR
600 0 8
500 6 0
400 5 4
300 1 0
4 rows × 2 columns
Awesome!
In [96]:
df2.plot()
Out[96]:
But when i Plot it looks the same (/sadpanda)
Any ideas =)?
回答1:
Another approach would be to invert the direction of the x-axis in matplotlib. The key bit of code here is plt.gca().invert_xaxis(). Note: this leaves the x-axis as an integer axis. Example code follows:
from StringIO import StringIO # for python 2.7; import from io for python 3
import pandas as pd
import matplotlib.pyplot as plt
# get data
data = """,sample1, sample2
300, 1, 0
400, 5, 4
500, 6, 0
600, 0, 8"""
df = pd.read_csv(StringIO(data), header=0, index_col=0, skipinitialspace=True)
# and plot
df.plot()
plt.gca().invert_xaxis()
plt.show()
回答2:
Make the index a string index - then plot will treat it as categorical data, and plot in the order it appears in the dataframe.
from StringIO import StringIO # for python 2.7; import from io for python 3
import pandas as pd
import matplotlib.pyplot as plt
data = """,sample1, sample2
300, 1, 0
400, 5, 4
500, 6, 0
600, 0, 8"""
df = pd.read_csv(StringIO(data), header=0, index_col=0, skipinitialspace=True)
# string index
df.index = df.index.astype(str)
# reverse dataframe
df = df[::-1]
# plot
df.plot()
plt.show()
来源：https://stackoverflow.com/questions/29511645/pandas-plot-x-or-index-column-in-descending-order