Trying to subplot groups with date axis different methods - pandas

Question

I got a DataFrame:
date phone sensor pallet
126 2019-04-15 940203 C0382C391A4D 47
127 2019-04-15 940203 C0382D392A4D 47
133 2019-04-16 940203 C0382C391A4D 47
134 2019-04-16 940203 C0382D392A4D 47
138 2019-04-17 940203 C0382C391A4D 47
139 2019-04-17 940203 C0382D392A4D 47
144 2019-04-18 940203 C0382C391A4D 47
145 2019-04-18 940203 C0382D392A4D 47
156 2019-04-19 940203 C0382D392A4D 47
157 2019-04-19 940203 C0382C391A4D 47
277 2019-04-15 941557 C0392D362735 32
279 2019-04-15 941557 C03633364D50 32
286 2019-04-16 941557 C03633364D50 32
287 2019-04-16 941557 C0392D362735 32
296 2019-04-17 941557 C03633364D50 32
297 2019-04-17 941557 C0392D362735 32
305 2019-04-18 941557 C0392D362735 32
306 2019-04-18 941557 C03633364D50 32
317 2019-04-19 941557 C03633364D50 32
318 2019-04-19 941557 C0392D362735 32
561 2019-04-15 942316 C0384639224D 45
562 2019-04-15 942316 C03632364950 45
563 2019-04-15 942316 C03920363835 45
564 2019-04-15 942316 C0382939384D 45
573 2019-04-16 942316 C0382939384D 45
574 2019-04-16 942316 C0384639224D 45
575 2019-04-16 942316 C03632364950 45
i want to be able to make subplot for each pallet which contain the sensors arrived in each date.
example:
i have tried few methods:
ax.plot_date
looping through opened ax's and plotting through each 1
grouped = pallets_arrived.groupby('pallet')
nrows = 2
ncols = 2
fig, axs = plt.subplots(nrows, ncols)
targets = zip(grouped.groups.keys(), axs.flatten())
for i, (key, ax) in enumerate(targets):
ax.plot_date(grouped.get_group(key)['date'], grouped.get_group(key)['sensor'], 'o')
plt.show()
return pallets_arrived
which gives wierdly formatted repeating dates (index the Df with date isnt solving the prob)
Df plotting
grouped = pallets_arrived.groupby('pallet')
nrows = 2
ncols = 2
fig, axs = plt.subplots(nrows, ncols)
targets = zip(grouped.groups.keys(), axs.flatten())
for i, (key, ax) in enumerate(targets):
grouped.get_group(key).plot(x='date', y='sensor', ax=ax)
ax.legend()
plt.show()
or
grouped = pallets_arrived.set_index('date').groupby('pallet')
nrows = 2
ncols = 2
fig, axs = plt.subplots(nrows, ncols)
targets = zip(grouped.groups.keys(), axs.flatten())
for i, (key, ax) in enumerate(targets):
grouped.get_group(key).plot(grouped.get_group(key).index, y='sensor', ax=ax)
ax.legend()
plt.show()
pyplot
grouped = pallets_arrived.groupby('pallet')
nrows = 2
ncols = 2
fig, axs = plt.subplots(nrows, ncols)
targets = zip(grouped.groups.keys(), axs.flatten())
for i, (key, ax) in enumerate(targets):
plt.sca(ax)
plt.plot(grouped.get_group(key)['date'], grouped.get_group(key)['sensor'])
ax.legend()
plt.show()
which again
Pivot pallets to Plot() on columns(pallets)
which doesnt work because there are more than 1 sensor in each pallet in same date. so there is a duplicated value error...
I really dont know what method to use to make this 1 correct:
grouping similar dates in x axis.
being able to plot each pallet to different subplot.
i think i dont get the pandas wrapping of matplotlib correctly.
ill be glad for some explenation because im reading guides and cant understand the preferred method for those stuff.
Thanks alot for the helpers.

Answer 1

you can use matplotlib to plot categorical data:
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib import style
style.use('ggplot')
%matplotlib inline
fig, ax = plt.subplots()
ax.scatter(df['date'], df['sensor'])
plt.show()
or if you want to color the groups:
fig, ax = plt.subplots()
for _,g in df.groupby('pallet'):
ax.scatter(g['date'], g['sensor'])
plt.show()
you can also add a legend:
fig, ax = plt.subplots()
for _,g in df.groupby('pallet'):
ax.scatter(g['date'], g['sensor'], label='Pallet_'+str(_))
ax.legend()
plt.show()