Do EDA on open-meteo's weather data #1

[CarlosGDCJ]

Try to get a better sense of the data by creating a notebook with visualizations. Open-meteo data is located in the folder data/weather in a file called open_meteo.csv. It contains historical weather data since 2010 from 10 cities all around Myanmar.

You can check the meaning of all variables in the picture below:

The visualization notebook you create should be placed in the folder src/weather/visualizations.

[FayElhassan]

Here what I was able to do so far please let me know if this is enough or needs something more https://colab.research.google.com/drive/1MvmgtHBHKQwjArFtGKjoXa64eVmbrny-?usp=sharing

[CarlosGDCJ]

@FayElhassan You came up with some nice visualizations, but you didn't write the key insights you gained from the graphs you created. I liked that you put titles and labeled the axis in your plots with relevant names and the unit of measure; but the goal of EDA is to discover characteristics of the data that could be helpful in solving the problem. Writing your findings is extremely important.

Some plots are confusing(I don't know what you're trying to do with them). For example, this one: This plot clearly has a problem: The same color is attributed to multiple stuff and the legend is so big it goes off-screen.

You had other formatting problems like this plot: Change the x ticks so everything doesn't get clumped up together.

Tips:

• Make sure every picture in your EDA has a reason to be there: some people like to do EDA in a question-answer format, where you come up with a question and create a visualization that answers said question. That gives every plot a clear justification, and at the same time by answering the question you communicate your findings.
• You're using seaborn but you're getting matplotlib styles (I don't know if that was your intention). If you want the seaborn styles, you need to call sns.set()
• If your axis shows only the year (2010, 2012, ...) it's better to label it as "Year" instead of "Date"
• We could add data from different SEA countries later on. To make sure your EDA doesn't break, filter your data for Myanmar only with something like myanmar_data = data[data["country"] == "Myanmar"]
• Weather tends to vary as months pass (this can be seen in some of the plots). An analysis explicitly showing this would be great.

[FayElhassan]

Okay very much noted I will look into it again