How to plot multiple dataframes with different lenghts into one plot

You can do something like this

import numpy as np
import matplotlib.pyplot as plt

df1 = [5, 3, 10, 7, 10]
df2 = [2, 4, 5, 7, 2, 1, 3, 0, 1, 3]

arr1 = []
for i in range(len(df1)):
    arr1.append([i,df1[i]])
    
arr1 = np.array(arr1, dtype='int')
x1 = arr1[:,0]
y1 = arr1[:,1]

arr2 = []
for j in range(len(df2)):
    arr2.append([j,df2[j]])
    
arr2 = np.array(arr2, dtype='int')
x2 = arr2[:,0]
y2 = arr2[:,1]

fig = plt.figure(figsize=(9,4))
ax1 = fig.add_subplot(111)
ax1.plot(x1,y1, color='red')
ax1.plot(x2,y2, color='blue')
plt.xlabel('Index')
ax1.legend(["Serie A", "Serie B"], loc ="upper right")
plt.title('Seria A and Serie B', loc='left', color='grey')
plt.grid()
plt.show()

In case you just want to make them same size:

If you already have pandas dataframes:

import pandas as pd
import math

# generate sample data
adf = pd.DataFrame({"df1" : [5,3,10,7,10,5,8,6,5,6,2,1,5,3,6,9,9,9,7,10,2,7,7,6,1,6,9]})
bdf = pd.DataFrame({"df2" : [2,4,6,7,2,1,3,0,1]})

# add the bdf values to adf dataframe
adf['bdf'] = [bdf['df2'].iloc[math.floor(i/len(adf.df1)*len(bdf.df2))] for i in range(len(adf.df1))]

adf
#Out[3]: 
#    df1  bdf
#0     5    2
#1     3    2
#2    10    2
#3     7    4
#4    10    4
#5     5    4
#6     8    6
#7     6    6
#8     5    6
#9     6    7
#10    2    7
#11    1    7
#12    5    2
#13    3    2
#14    6    2
#15    9    1
#16    9    1
#17    9    1
#18    7    3
#19   10    3
#20    2    3
#21    7    0
#22    7    0
#23    6    0
#24    1    1
#25    6    1
#26    9    1

If you have just a list of values:

a = [5,3,10,7,10,5,8,6,5,6,2,1,5,3,6,9,9,9,7,10,2,7,7,6,1,6,9]
b = [2,4,6,7,2,1,3,0,1]
b_adjusted = [b[math.floor(i/len(a)*len(b))] for i in range(len(a))]
b_adjusted
#Out[5]: 
#[2,
# 2,
# 2,
# 4,
# 4,
# 4,
# 6,
# 6,
# 6,
# 7,
# 7,
# 7,
# 2,
# 2,
# 2,
# 1,
# 1,
# 1,
# 3,
# 3,
# 3,
# 0,
# 0,
# 0,
# 1,
# 1,
# 1]

Updated Answer

After you’ve clarified the problem in your question and comments, you’re actual intention is not to combine the datasets but to simply plot the datasets such that the first and last points match up. This can easily be done by creating an x array for each data frame that goes from 0 to 1 but has the N points, where N is the length of the data frame; I did this using np.linspace.

Shown below is the code and plots for the original and time series data. Since you want to see the trends, for the time series data I subtracted each of them by the first value so they all start at 0.

Using the original data:

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

plt.close("all")

df1 = pd.DataFrame({"data": [5, 3, 10, 7, 10, 5, 8, 6, 5, 6, 2, 1, 5, 3, 6, 9, 9, 9, 7, 10, 2, 7, 7, 6, 1, 6, 9]})
df2 = pd.DataFrame({"data": [2, 4, 6, 7, 2, 1, 3, 0, 1]})

fig, ax = plt.subplots()
for i, df in enumerate([df1, df2], start=1):
    x = np.linspace(0, 1, len(df))
    ax.plot(x, df["data"], label=f"df{i}")
ax.legend()
fig.show()

Using the time-series data:

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

plt.close("all")

df1 = pd.DataFrame({"datetime": ["2020-01-19 22:05:00",
                                 "2020-01-19 22:10:00",
                                 "2020-01-19 22:15:00",
                                 "2020-01-19 22:20:00",
                                 "2020-01-19 22:25:00",
                                 "2020-01-19 22:30:00",
                                 "2020-01-19 22:35:00",
                                 "2020-01-19 22:40:00",
                                 "2020-01-19 22:45:00",
                                 "2020-01-10 22:50:00",],
                    "value": [166.300003,
                              165.259995,
                              164.699997,
                              165.380005,
                              166.179993,
                              162.630005,
                              162.550003,
                              161.990005,
                              161.750000,
                              161.440002,
                              ]})
df2 = pd.DataFrame({"datetime": ["2020-02-25 11:55:00",
                                 "2020-02-25 12:00:00",
                                 "2020-02-25 12:05:00",
                                 "2020-02-25 12:10:00",
                                 "2020-02-25 12:15:00",
                                 "2020-02-25 12:20:00",
                                 "2020-02-25 12:25:00",
                                 "2020-02-25 12:30:00",
                                 "2020-02-25 12:35:00",
                                 "2020-02-25 12:40:00",
                                 "2020-02-25 12:45:00",
                                 "2020-02-25 12:50:00",
                                 "2020-02-25 12:55:00",
                                 "2020-02-25 13:00:00",
                                 "2020-02-25 13:05:00",
                                 "2020-02-25 13:10:00",
                                 "2020-02-25 13:15:00",],
                    "value": [262.510010,
                              263.179993,
                              262.260010,
                              261.959991,
                              262.179993,
                              261.299988,
                              261.579987,
                              261.890015,
                              262.820007,
                              262.010010,
                              261.630005,
                              261.109985,
                              261.149994,
                              260.679993,
                              261.929993,
                              260.880005,
                              259.929993,]})
df3 = pd.DataFrame({"datetime": ["2020-02-27 06:10:00",
                                 "2020-02-27 06:15:00",
                                 "2020-02-27 06:20:00",
                                 "2020-02-27 06:25:00",
                                 "2020-02-27 06:30:00",],
                    "value": [224.860001,
                              224.240005,
                              223.610001,
                              223.490005,
                              223.199997,]})

fig, ax = plt.subplots()
for i, df in enumerate([df1, df2, df3], start=1):
    x = np.linspace(0, 1, len(df))
    ax.plot(x, df["value"] - df["value"].iloc[0], label=f"df{i}")
ax.legend()
fig.show()

---

Original Answer

Although I think this is not the right way to deal with your data, I will show you how to do what you ask.

Before we implement any of the methods, you will need to combine the data with the alignment you showed. To do this, you can reindex df2 in steps of len(df1)/len(df2). I do this using the following code:

import pandas as pd
import numpy as np

df1 = pd.DataFrame({"df1": [5, 3, 10, 7, 10, 5, 8, 6, 5, 6, 2, 1, 5, 3, 6, 9, 9, 9, 7, 10, 2, 7, 7, 6, 1, 6, 9]})
df2 = pd.DataFrame({"df2": [2, 4, 6, 7, 2, 1, 3, 0, 1]})
df2.index = np.arange(0, len(df1), len(df1)//len(df2))

The two DataFrames can then be joined on the index column using this next line.

df_combined = df1.join(df2, how="outer")

So, after those two steps, we have df_combined.

    df1  df2
0     5  2.0
1     3  NaN
2    10  NaN
3     7  4.0
4    10  NaN
5     5  NaN
6     8  6.0
7     6  NaN
8     5  NaN
9     6  7.0
10    2  NaN
11    1  NaN
12    5  2.0
13    3  NaN
14    6  NaN
15    9  1.0
16    9  NaN
17    9  NaN
18    7  3.0
19   10  NaN
20    2  NaN
21    7  0.0
22    7  NaN
23    6  NaN
24    1  1.0
25    6  NaN
26    9  NaN

Note: Because NaNs were added, the df2 column is now dtype=float.

You seem to mention two methods, filling df2 with the data with the same value and interpolating. For filling, you can simply call the ffill() method, which will forward-fill NaNs, i.e. if the value is 2 before a NaN, it will fill that in with a 2.

df_combined["df2_ffill"] = df_combined["df2"].ffill()

To interpolate, you can call the interpolate() method, which, by default, will linearly interpolate the NaNs based on the values before and after the NaN (or group of NaNs).

df_combined["df2_interpolate"] = df_combined["df2"].interpolate()

So, we have df_combined:

    df1  df2  df2_ffill  df2_interpolate
0     5  2.0        2.0         2.000000
1     3  NaN        2.0         2.666667
2    10  NaN        2.0         3.333333
3     7  4.0        4.0         4.000000
4    10  NaN        4.0         4.666667
5     5  NaN        4.0         5.333333
6     8  6.0        6.0         6.000000
7     6  NaN        6.0         6.333333
8     5  NaN        6.0         6.666667
9     6  7.0        7.0         7.000000
10    2  NaN        7.0         5.333333
11    1  NaN        7.0         3.666667
12    5  2.0        2.0         2.000000
13    3  NaN        2.0         1.666667
14    6  NaN        2.0         1.333333
15    9  1.0        1.0         1.000000
16    9  NaN        1.0         1.666667
17    9  NaN        1.0         2.333333
18    7  3.0        3.0         3.000000
19   10  NaN        3.0         2.000000
20    2  NaN        3.0         1.000000
21    7  0.0        0.0         0.000000
22    7  NaN        0.0         0.333333
23    6  NaN        0.0         0.666667
24    1  1.0        1.0         1.000000
25    6  NaN        1.0         1.000000
26    9  NaN        1.0         1.000000

Here is the data plotted:

Now, as I said in the beginning, this is not the best way to deal with the data because there is no "right" way to make up data. The best thing to do is what you mention last, which is to plot against some other variable; the DataFrame index is not a good option to plot against because it has no meaning.

Assuming this data is some dependent data, i.e. it was computed or derived by some process, there is probably an independent variable associated with it. Maybe it’s time, such as seconds, days, or years, or it’s distance, such as feet, meters, or miles. That is what you should actually plot against. If you have that information for df1 and df2, then plotting each against the corresponding independent data will line up the graphs properly and you wouldn’t have to combine them and make up data.