# Practice Problems¶

### Lecture 17¶

Answer each number in a separate cell

Rename this notebook with your last name and the lecture

ex. CychB_17



Turn-in this notebook on triton-ed by the end of class

# 1. Kernel Density Estimates and Jointplots¶

• Use pandas to read in the tab separated file Datasets/ZirconData.txt. These data are from Bisnath et al (2006) and were downloaded from the Pangaea database (https://doi.pangaea.de/10.1594/PANGAEA.847015)
• These data are Lead-Lead and Uranium-Lead age dates for zircons from Dronning Maud Land, Antartica. The zircon ages show two stages of growth associated with two stages of mountain building in the region.

• The columns ['Age dated [ka] (Age, 206Pb/238U Lead-Uranium)'] and ['Age dated [ka] (Age, 207Pb/206Pb Lead-Lead)'] are the different age systems. Get the mean of each of these ages.

• Use plt.hist() to plot a histogram of both age dates.
• Use sns.kdeplot() to plot a kernel density of both age systems. Experiment with the bw argument to see how this alters your result.
• Use sns.jointplot() to plot a 2d kernel density estimate of the data.

# 2. Box and Violin Plots¶

• Plot a boxplot of the Pb-U age for each sample (designated by the 'Sample ID' column). Plot the mean Pb-U age for the whole dataset as a horizontal line.
• Plot a violin plot for each of these samples. Plot a swarm plot on top of this plot.
• Does the boxplot or the violin plot tell you more about the data in this case?

# 3. Pairplots¶

• Filter the DataFrame to only include rows with 207Pb/206Pb and 206Pb/238U <0.5
• Plot the four elemental ratios 232Th/238U, 207Pb/206Pb, 207Pb/235U and 206Pb/238U against one another
• Can you see a bimodal distribution in any of these plots?
• Are any of these ratios strongly correlated?