CODEWITHSUNDEEP
pythonpandasmatplotlibdataframe
Lucas Amos
Lucas Amos

Reputation: 1195

matplotlib pyplot not accurately representing time along x axis

I have a JSON array of around 3000 records each spaced 1 minute apart which I want to plot on a timeseries graph using matplotlib. However I cannot get the data to display properly on a chart with the date and times along the x axis.

The records are formatted as follows:

{
    "datetime": "10-07-2017  21:57:37", 
    "deviceid": "PiJCLabDoor", 
    "dht11": 24.0, 
    "dht22": null, 
    "id": 39751, 
    "motion": 0.0, 
    "pressure": null, 
    "uv": 0.04, 
    "voc": 126.0
  }



I have the following code

import pandas as pd
from pandas import DataFrame
import matplotlib.pyplot as plt

data = pd.read_json('PiJCLabDoor.json', convert_dates=['datetime'])
dataframe = DataFrame(data)

for i in range(0, dataframe.shape[0] -1):

    if (dataframe.loc[i, 'dht11'] - dataframe.loc[i+1, 'dht11']) > 3 :
    dataframe.loc[i+1, 'dht11'] = dataframe.loc[i,'dht11']

These lines of code yield the following chart:

plt.figure(figsize=(20,7)
plt.plot(dataframe['datetime'], dataframe['dht11'])
plt.show()

enter image description here

And these lines of code yield the following chart:

plt.rcParams['figure.figsize'] = (20,6)
plt.plot(dataframe['dht11'])
plt.show()

enter image description here

The second chart is what the data should look like however the first chart has dates on the x axis.

How do I get the date and time to appear on the x axis of the second chart? Thanks

Upvotes: 0

Views: 1732

Answers (1)

Kacper Wolkowski
Kacper Wolkowski

Reputation: 1597

First thing, remember to convert datetime to Pandas datetime (pd.to_datetime()) To make the dates appear on second graph set datetime column as index, something like:

plt.plot(dataframe.set_index('datetime')['dht11'])

EDIT:

So if i get it right you have something like this:

enter image description here

There is a 1 week gap between dates, and you want to have them one next to another so:

x = np.arange(len(a['date']))

plt.plot(x, a['value'])
plt.xticks(x, a['date'], rotation=45)

First I do something like in your first graph, then I change the name of xlabels to have the same as dates. Result looks like that:

enter image description here

But you might find that the xticks appears too often, then just change one line there, like this:

plt.xticks(x[::3], a['date'][::3], rotation=45)

So now only every third date will appear on the graph.

EDIT 2:

If you want to set only X main dates on your x-axis use this:

from matplotlib.ticker import LinearLocator
plt.plot(x, a['value'])
plt.xticks(x, a['date'], rotation=45)
ax = plt.gca()
ax.get_xaxis().set_major_locator(LinearLocator(numticks=5))

In numticks you specify how many ticks should appear

Upvotes: 3

Related Questions