Very strange plot from basic ASCII file data-CodePudding

Working with Jupyter Lab!

I loaded a simple ASCII file (as I've done 100s of times before...), with three columns and 2000 lines. I did it with

with open(file) as f: 
    d = f.readlines()

and also with NumPy d = np.loadtxt(file, delimiter=',') to see if something would change.

All values in the 2nd column (Latitude) are -32. and something and all values in the 3rd column (Longitude) are -52. and something. However, de variation of the data is about the 5th decimal number... (I think this is making things weird!)

When I printed the data on the screen, it seems ok! But when I try to plot them, I got pretty weird stuff... the numbers in the X and Y axis are nonsense, especially the one scaling the x-axis. 1e-5-5.2103e1

I open the data in a spreadsheet (Libre Calc), and there the plot seems alright. Then I save it with another name and tried to load in the Jupyter again and got the same weird result.

I also tried it also using a different computer... same result!

Tried a script using Atom... same result!

Can someone give a clue about what is going on?

The file is shared at:

CodePudding user response：

It looks like your plot is correct, even though the scaling of the axes is quite strange (likely due to the very small range). I plotted the same data using Altair, which was able to handle the axes ranges much better:

import pandas as pd
import altair as alt
with open("lat_long.csv", newline="") as f:
    frame = pd.read_csv(f, delimiter=",", header=None, names=["index", "longitude", "latitude"])

alt.Chart(frame).mark_circle(size=10).encode(
    alt.X('latitude',
        scale=alt.Scale(zero=False)
    ),
    alt.Y('longitude',
        scale=alt.Scale(zero=False)
    ),
).interactive()

The result matches yours, except for the more sensible axes:

My guess is that the pandas plotter is just doing something strange with displaying the axes, but the data is being read in correctly.

CodePudding user response：

This is how matplotlib displays very large and small tick markers. For example, to read the y-axis you take that number at the top and add each tick marker. I'll admit, the number on the x-axis is weird, but you can see a recognizable prefix in there: -5.2103e1. The 1e-5 might just be signifying the decimal place the ticks start at. Looking through the file, the plot makes sense, and the labels do as well.