在 Pandas / Matplotlib 中使用 GroupBy 函数绘制多条线

Question

I'm trying to plot a pandas dataframe using matplotlib however having issues with the grouping. The dataframe contains statistics for a player in each round of the season. My dataframe is much larger however for this example I have simplified it:

Desc    Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7 Round 8 Round 9 Round 10
Ben     22.3    33.3    21.5    27.7    31.3    43      33.5    20      29.7    22.7
Tom             28.2    29.2    23.1    25      21.4    22.3    26.2    25.3    19.6
Jack    21.3    30.4    20.8    18      24.5    28.3    32.6    17      25.1    23.7

However when I simply try to plot this using:

df.plot()
plt.show()

The lines are grouped by the round number instead of the player's name and it appears the Y values are actually the player's row index. Here is the plot it outputs.

So I believe maybe the pandas dataframe isn't corrected indexed for rows / columns thus causing this problem. I've looked into using the df.groupby but can't find a solution.

I can easily create the line graph I'm after using MS Excel - Here is the output I would like:

Does anyone have a solution on what I can do to either my dataframe or plot code to get the desired outcome? I have already made sure I have set the row Index's to the players name using:

df.set_index('Desc')

However this hasn't fixed the issue.

Answer 1

Use set_index then transpose:

Creating data

colNames = ['Desc', 'Round1', 'Round2', 'Round3', 'Round4', 'Round5', 'Round6', 'Round7', 'Round8', 'Round9', 'Round10']
df = pd.DataFrame(columns = colNames)
df.loc[len(df)] = ['Ben', '22.3', '33.3', '21.5', '27.7', '31.3', '43', '33.5', '20', '29.7', '22.7']
df.loc[len(df)] = ['Tom', '', '28.2', '29.2', '23.1', '25', '21.4', '22.3', '26.2', '25.3', '19.6']
df.loc[len(df)] = ['Jack', '21.3', '30.4', '20.8', '18', '24.5', '28.3', '32.6', '17', '25.1', '23.7']

Pre-processing

df.set_index("Desc", inplace = True)
df = df.apply(pd.to_numeric, errors='coerce')

Plotting the data

df.T.plot()
plt.show()

This gives us expected graph :