CODEWITHSUNDEEP
pythonpandasseaborn
Eran
Eran

Reputation: 844

seaborn scatterplot datetime xaxis too wide

I have this dataframe:

pd.DataFrame({'Depth': {0: 0.2,
  1: 0.4,
  2: 0.4,
  3: 0.4,
  4: 0.4,
  5: 0.4,
  6: 0.6000000000000001,
  7: 0.4,
  8: 3.2,
  9: 2.0},
 'DateTimeUTC': {0: Timestamp('2018-03-28 06:25:08'),
  1: Timestamp('2018-03-28 06:25:49'),
  2: Timestamp('2018-03-28 06:27:06'),
  3: Timestamp('2018-03-28 06:32:11'),
  4: Timestamp('2018-03-28 06:32:59'),
  5: Timestamp('2018-03-28 06:34:02'),
  6: Timestamp('2018-03-28 06:35:38'),
  7: Timestamp('2018-03-28 06:37:04'),
  8: Timestamp('2018-03-28 06:39:08'),
  9: Timestamp('2018-03-28 06:40:52')}})

which looks like this:

<table>
  <tr><th></th><th>Depth</th><th>DateTimeUTC</th></tr>
  <tr><th>0</th><td>0.2</td><td>2018-03-28 06:25:08</td></tr>
  <tr><th>1</th><td>0.4</td><td>2018-03-28 06:25:49</td></tr>
  <tr><th>2</th><td>0.4</td><td>2018-03-28 06:27:06</td></tr>
  <tr><th>3</th><td>0.4</td><td>2018-03-28 06:32:11</td></tr>
  <tr><th>4</th><td>0.4</td><td>2018-03-28 06:32:59</td></tr>
  <tr><th>5</th><td>0.4</td><td>2018-03-28 06:34:02</td></tr>
  <tr><th>6</th><td>0.6</td><td>2018-03-28 06:35:38</td></tr>
  <tr><th>7</th><td>0.4</td><td>2018-03-28 06:37:04</td></tr>
  <tr><th>8</th><td>3.2</td><td>2018-03-28 06:39:08</td></tr>
  <tr><th>9</th><td>2.0</td><td>2018-03-28 06:40:52</td></tr>
</table>

Note the all DateTimeUTC are within 2018. When I try to plot depth vs time using sns.scatterplot I get:

sns.scatterplot('DateTimeUTC', 'Depth', data=df)

plot

Why does the X-axis start at year 2000? Am I doing something wrong?

Upvotes: 9

Views: 4999

Answers (2)

Eran
Eran

Reputation: 844

I posted the question as an issue on Github, and got this great response. Basically, the problem is that plt.scatter does not deal with dates well, and seaborn uses it. If seaborn will add a type check for the x-axis, which uses plt.plot_date for date values instead, this will be fixed. In the meanwhile, one can create a custom version of sns.scatterplot that does excatly that.

Upvotes: 3

Sander van den Oord
Sander van den Oord

Reputation: 12808

As an alternative you could use seaborns lineplot which does have a correct x-axis:

sns.lineplot(x='DateTimeUTC', y='Depth', data=df, marker='o')

Or you could use:

plt.plot(df['DateTimeUTC'], df['Depth'], linestyle='None', marker='o')

Upvotes: 1

Related Questions